Introduction
Problem statement
Foreign direct investment (FDI) and exports play a crucial role in driving economic growth, particularly in developing countries, as highlighted by both theoretical and empirical literature New growth theories emphasize the significance of investment and trade openness, leading to strategies that promote FDI and export growth, often referred to as 'FDI-led growth' or 'export-led growth.' International free trade is recognized as an "engine of growth," with studies showing that countries embracing free trade experience higher growth rates compared to those that restrict trade (Sachs and Warner, 1997; Wacziarg and Welch, 2008) An open economy fosters greater economic growth, elevating income levels and living standards (Frankel and Romer, 1999) Through trade, export sectors can generate higher profits, increase savings, and create job opportunities (Azam, 2010) Economic output growth can be achieved not only by increasing labor and capital but also by expanding exports Additionally, competitive export markets encourage export-oriented policies that positively impact economic growth and facilitate integration into the global economy Thus, the export-led growth hypothesis supports the notion that export expansion is a key determinant of economic growth, while productivity improvements and cost reductions further enhance export capabilities (Krugman, 1984).
Foreign direct investment (FDI) is on the rise, playing a crucial role in enhancing trade and economic growth It boosts productivity, generates employment, advances technology, improves social welfare, and alleviates poverty, particularly in developing countries Research by Chiara and Sasidharan (2009) highlights that FDI inflows are vital for economic development, as they introduce unique technological assets unavailable in the host country Additionally, trade serves as a conduit for FDI, facilitating the transfer of capital, technological advancements, and managerial expertise.
Foreign direct investment (FDI) significantly enhances economic growth and plays a crucial role in promoting exports Consequently, the interplay between FDI, exports, and economic growth has become increasingly important to policymakers and researchers alike.
Economists and researchers have long been interested in the relationship between Foreign Direct Investment (FDI), exports, and economic growth Numerous empirical studies have utilized time series, cross-section, and panel data to explore the impact of FDI and exports on economic growth, focusing on the FDI-led and export-led growth hypotheses Early research employed simple correlation coefficients and regression equations based on Neoclassical growth models, with exports as an explanatory variable Recent studies have shifted towards emphasizing causality between FDI, export growth, and economic growth through cointegration techniques and Granger causality analysis Various empirical studies have applied these methods to investigate the causal links among FDI, exports, and economic growth in regions such as Asia and Africa, particularly in developing countries, with notable examples including Ekanayake (1999) and Ismaid and Harjito (2003).
Numerous studies, including those by Sinoha-Lopete (2006), Pham (2008), and Awan et al (2012), highlight the complex relationship between Foreign Direct Investment (FDI), exports, and economic growth, which varies by country and region It is crucial for policymakers and researchers to understand how these variables interact to formulate effective strategies that promote economic growth, particularly in developing nations Notably, Malaysia and Thailand exemplify developing countries that have achieved sustained economic growth over the long term.
Malaysia and Thailand have successfully boosted their economic growth through increased exports and foreign direct investment (FDI), particularly in the manufacturing sector, as noted by Todaro and Smith (2003) Vietnam shares similar economic conditions and comparative advantages, making the growth-promoting policies of these countries valuable lessons for national development This thesis investigates the causal relationships between FDI, exports, and economic growth using cointegration and causality techniques on time series and panel data from 1989 to 2010, focusing on Malaysia, Thailand, and Vietnam in the Southeast Asia region, which includes many developing nations The findings aim to enhance the empirical literature with modern methodologies.
Research objectives
The objectives are as follows:
This study investigates the causal relationship between foreign direct investment (FDI), exports, and economic growth in specific countries The analysis aims to understand how these factors interact and influence each other, providing insights into their economic dynamics.
• To suggest appropriate measures for exports expansion and attracting FDI in order to stimulate economic growth in these countries.
Research questions
In order to achieve the research objectives:
• Is there a causal relationship between FDI, exports, and economic growth in Malaysia, Thailand, and Vietnam countries?
• What policy implications could help improve exports and environment for FDI in these countries?
Research methodology
This paper uses annual time series data and panel data of three countries in Southeast Asia region including Malaysia, Thailand, and Vietnam over the period of 1989-2010
This study investigates the relationship between foreign direct investment (FDI), exports, and economic growth using various statistical methods For time series data, the analysis employs the Augmented Dickey-Fuller (ADF) test, Phillips and Perron (PP) test, Johansen's cointegration technique, and Granger causality test In contrast, the panel data analysis utilizes the Im, Pesaran, and Shin (IPS) test, Augmented Dickey-Fuller Fisher (ADF-Fisher) test, Johansen Fisher panel cointegration, Kao tests, and panel Granger causality test.
Structure of the thesis
This thesis comprises six chapters, beginning with Chapter 1, which outlines the problem definition, objectives, research questions, and a brief overview of the methodology Chapter 2 offers a comprehensive theoretical and literature review In Chapter 3, an overview of the issues surrounding foreign direct investment, exports, and economic growth is presented Finally, Chapter 4 details the methodology employed in the study.
Chapter 5 indicates empirical results corresponding to each estimation technique
The final chapter presents the conclusions drawn from empirical findings, discusses the implications for policy, and addresses the limitations of the thesis.
Literature Review
Theoretical Literature
2.1.1 Relationship between foreign direct investment and economic growth
Foreign direct investment (FDI) serves as a crucial capital source, enhancing domestic private investment and creating new job opportunities Additionally, FDI facilitates technology transfer, which significantly contributes to economic growth in host countries Various mechanisms exist through which FDI can drive this economic expansion.
The Neoclassical growth model views technological progress and labor force as exogenous factors, suggesting that foreign direct investment (FDI) boosts investment volume and economic growth, albeit with only a short-term effect due to diminishing returns to capital In contrast, the new endogenous growth theory posits that long-term growth is driven by technological change, indicating that if FDI positively impacts technology, it can enhance growth rates in host economies through technology transfer and spillover effects Research by Herzer et al (2006) confirms that FDI positively influences economic growth in both the short and long run, while Makki and Somwaru (2004) highlight the significance of FDI in facilitating technology transfer from developed to developing countries.
Economic growth significantly influences foreign direct investment (FDI) inflows into a host country A higher level of economic growth can attract FDI by expanding domestic markets and fostering business development (Agiomirgianakis et al., 2006).
Rapid economic growth significantly boosts aggregate demand, which in turn stimulates an increase in investments, including foreign direct investment (FDI) Additionally, improved economic performance in host countries enhances the attractiveness for FDI by providing better infrastructure and greater profit-making opportunities.
Johnson (2006) state that FDI inflows enhance economic growth of a host country and the growth of economic in tum may attract FDI inflows and the cycle continues
2.1.2 Relationship between exports and economic growth
Exports play a crucial role in driving economic growth, particularly in developing countries, as they are often seen as interconnected The export-led growth hypothesis is supported by various theoretical arguments that highlight the importance of exports in fostering economic development.
Early research by Balassa (1978) indicates a positive relationship between export growth and economic development, highlighting a significant correlation between the two This study also finds a positive correlation between exports and domestic savings, suggesting that an increase in exports can attract substantial foreign capital and enhance labor growth.
This paper supports the notion that export-oriented policies drive economic growth Tyler (1981), aligning with Balassa, analyzed the relationship between growth and export expansion using cross-sectional data from 55 developing countries, revealing a strong positive correlation between the two However, this correlation analysis does not account for the influence of other variables To address this, he developed a model based on the Cobb-Douglas production function, incorporating three key factors.
Where Xi is country i's GNP, A is a technological constant, Ki is country i's capital stock services, Li is a country i' s labor force inputs and Ei is country i' s exports
This regression analysis highlights the significant impact of capital and exports on economic growth, surpassing models that only consider capital and labor The findings demonstrate that a 17.5% increase in exports correlates with a 1% increase in growth, underscoring the crucial role of exports in driving economic development in developing countries.
Export orientation and promotion are advantageous for both developed and developing nations, as they leverage economies of scale and technological advancements to boost productivity and create jobs Expanding exports alleviates current account pressures by enhancing external earnings and attracting foreign investment Additionally, promoting export activities fosters economic growth by encouraging the production of goods for international markets, leading to specialization and the utilization of a nation's comparative advantages.
Export expansion is crucial for economic growth, as it increases foreign demand for domestic products, leading to job creation and higher incomes in the export sector This growth in exports facilitates the efficient reallocation of resources and the realization of scale economies, resulting in enhanced productivity Additionally, increased exports drive advancements in technology, skill development, and management practices due to competition in foreign markets This influx of foreign exchange allows for greater imports of capital and intermediate goods, further boosting output growth The relationship between exports and output growth is reciprocal; as exports rise, they generate more employment and income, which in turn fosters productivity gains and stimulates further export growth.
Besides, there are also theoretical reasons to support the growth-led export
Neoclassical trade theory posits that economic growth fosters higher exports, primarily through economies of scale that reduce costs and enhance comparative advantage As economies grow, they also improve skills and technology across various sectors, leading to increased productivity (Krugman, 1984) Similarly, Ram (2003) emphasizes that advancements in skills and technology provide a comparative advantage in producing certain goods, thereby expanding exports Furthermore, Majeed and Ahmad (2006) demonstrate that GDP and GDP growth positively impact exports, particularly in developing countries.
Therefore, it is important to maintain a high and sustainable economic growth which is a vital factor to promote exports because bidirectional causality between exports and output growth can exist
2.1.3 Relationship between FDI and exports
Foreign Direct Investment (FDI) is primarily driven by the potential for high profits in emerging markets It influences the export performance of host countries through the activities of foreign affiliates and the increased exports of local firms (Lipsey, 2004) Multinational corporations (MNCs) leverage their affiliates' resource advantages, such as abundant resources and lower labor costs, to enhance their competitiveness in global markets, thereby improving overall export performance.
Foreign Direct Investment (FDI) indirectly influences local firms' exports through several channels, enhancing their competitiveness Multinational corporations (MNCs) can transfer technology, management practices, and skills, which boosts production efficiency Additionally, Njong (2008) highlights that the shift of FDI from countries with higher labor costs to those with lower labor costs improves MNCs' export competitiveness and overall export performance.
Exports influence Foreign Direct Investment (FDI) through various channels Key factors driving FDI inflows include high income per capita, a strong orientation towards international trade, infrastructure development, and attractive rates of return on investment, as noted by Akinkugbe (2003) Additionally, Lim (2004) emphasizes the importance of market size, infrastructure quality, economic stability, and the presence of free trade zones Investment decisions are further shaped by fiscal incentives, the overall investment environment, labor costs, and trade openness.
Empirical studies
Babalola et al (2012) examined the interplay between exports, foreign direct investment (FDI), and economic growth in Nigeria from 1960 to 2009 Their findings revealed six cointegration vectors among FDI, capital formation, degree of openness, imports, and terms of trade, as determined by the Johansen cointegration test The study concluded that these variables are interconnected in the long run and significantly influence Nigeria's economy.
Javed et al (2012) investigated the interplay between foreign direct investment (FDI), trade, and economic growth in India, Bangladesh, Sri Lanka, and Pakistan from 1973 to 2010 using the Generalized Method of Moments (GMM) Their findings indicate that exports positively influence economic growth in all countries studied, as does FDI, with the exception of Sri Lanka Additionally, domestic investment and the labor force also contribute positively to economic growth.
Mangir (2012) used cointegration and Granger causality test to analyze the relationship between export and economic growth in Turkey for the period of 2002
The analysis of quarterly time series data from 2011 reveals a long-run cointegration between exports and economic growth in Turkey The Granger causality test indicates a unidirectional causality from exports to economic growth in the short run, while a bidirectional causality relationship is observed in the long run.
Hossain (2012) analyzed both the short-run and long-run relationship between FDI and economic output in South Asian countries using data from 1972 to
In a 2008 study utilizing three econometric models (ADF test, Engle-Granger test, and VECM) alongside Granger causality analysis, the impact of Foreign Direct Investment (FDI) on economic output was examined for Bangladesh, India, and Pakistan The findings revealed no cointegration between FDI and GDP in both the short and long run for Bangladesh and India However, Pakistan exhibited cointegration in both time frames Additionally, there was no causality relationship between FDI and GDP in Bangladesh, while a one-way relationship was identified in Pakistan and India, indicating that FDI influences economic output in these countries.
A study by Sun (2011) analyzed the relationship between foreign direct investment (FDI) and economic growth in China from 1985 to 2010 using a cointegration approach The findings indicate that FDI and economic growth are cointegrated, and the Granger causality test reveals that economic growth in China drives an increase in FDI.
Erecakar (2011) examined the interplay between growth, foreign direct investment (FDI), trade, and inflation in Turkey from 1970 to 2008 The study employed a cointegration test to identify long-term relationships among these variables, revealing the existence of a single cointegration vector The findings demonstrated that FDI, inflation, and trade surplus positively influence economic growth.
Tekin (2011) investigates the Granger causality between real GDP, real exports, and inward foreign direct investment (FDI) in Least Developed Countries from 1990 to 2009 Utilizing a novel panel-data methodology that incorporates seemingly unrelated regression (SUR) systems and Wald tests, the study identifies significant causal relationships among these variables The findings indicate support for export-led growth and FDI-led growth, while also revealing an inverse relationship between FDI and exports.
Safdari et al (2011) examined the causal relationship between exports and economic growth in Asian developing countries from 1988 to 2008 The study utilized panel data from thirteen countries, measuring GDP and exports in constant 2000 US dollars, with a GDP deflator applied To analyze the data, the research employed four panel unit root tests, including those developed by Levin et al (LLC, 2002), Im et al (IPS, 2003), and Breitung.
The study employs the Fisher-type test by Maddala and Wu (1999) and Choi (2001) to assess stationarity Additionally, it utilizes the panel cointegration tests proposed by Pedroni (1999, 2004) and Kao (1999) to examine the cointegration relationship between export and output growth.
This study employs a panel-VECM causality analysis using the Wald test to investigate the causal relationship between real exports and real GDP The findings indicate that exports do not Granger-cause economic growth; instead, there is a one-way causality from economic growth to exports in thirteen selected Asian developing countries.
Eusuf and Ahmed (2010) investigated the causal relationship between export and economic growth in South Asian countries using the Engle-Granger error correction model Their study analyzed annual time series data from seven countries, covering varying time periods: India, Nepal, Sri Lanka, and Pakistan from 1965 to 2005, and Bangladesh and the Maldives from 1980 to 2005.
This study analyzes the relationship between exports and economic growth in South Asian countries, utilizing indices such as the consumer price index, unit value index for exports, and GDP deflator The ADF and PP tests assess stationarity, while the Engle-Granger procedure examines long-run relationships Causality is evaluated through the F-test and error correction term Findings reveal that real exports and real GDP are cointegrated in Bangladesh, Pakistan, and Nepal Export-led growth is confirmed for Pakistan, Sri Lanka, and Bhutan in both the short and long term, whereas growth-led exports are observed in India, Nepal, and the Maldives However, no feedback export-led growth relationship is established for Bangladesh, leading to mixed results that do not conclusively support export-led growth across South Asian nations.
Bahmani-Oskooee and Economidou (2009) examined the relationship between export-led growth and growth-led exports using annual data from 61 least developed countries (LDCs) spanning 1960-1999 The study analyzed five key variables: gross domestic product, gross capital formation, exports, imports (in constant 1995 US dollars), and total labor force Employing Johansen's cointegration technique, the findings revealed that 14 countries exhibited no cointegration vectors, while the remaining countries demonstrated at least one cointegrating vector among the variables The relationship between exports and output was categorized into four groups: the first group showed a long-run feedback relationship for countries like Algeria and Ghana; the second group, including Burkina Faso and India, indicated no long-run relationship; the third group, comprising Congo and South Africa, found that increasing exports stimulated output; and the final group, including Benin and Colombia, exhibited a one-way relationship from output growth to export growth Overall, the results highlight the country-specific nature of these relationships.
Rudra et al (2009) conducted a study on the relationship between foreign direct investment (FDI) and economic growth in five ASEAN countries—Indonesia, Malaysia, Singapore, the Philippines, and Thailand—covering the period from 1970 to 2007 The findings reveal a long-run cointegration between FDI and economic growth at the panel level for all five countries, while individual analysis showed this relationship only for Singapore and Thailand Additionally, bi-directional causality was identified between the two variables at both levels for all countries, except for Malaysia, as determined by Granger causality.
Miankhel et al (2009) investigated the causal relationship between foreign direct investment (FDI), exports, and GDP in six emerging countries—India, Pakistan, Malaysia, Thailand, Chile, and Mexico—covering the period from 1970 to 2005, analyzing both short-run and long-run effects.
Conceptual framework
The conceptual framework illustrated in Figure 2.1 highlights the interconnections between foreign direct investment, exports, and economic growth It also outlines the research methodologies employed, such as time series and panel data analysis, to investigate the causal relationships among these variables This framework provides a comprehensive summary of the study's contents.
Fi~un: 2.1: Conceptual fnmu.•nmãk in this stud~
Capital sources Increasing domestic Cost reduction Create employment markets and
Job opportunities businesses Advancement Improving income skills and Technology transfer Good infrastructure technology Technological progress
• IPS and ADF Fisher tests
• Kao and Panel Johansen cointegration tests
[' F-D-1 -~) ~ [ _ _ E_x_po_rt_s -~) tot nghiep do wn load thyj uyi pl aluan van full moi nhat z z vbhtj mk gmail.com Luan van retey thac si cdeg jg hg
Foreign Direct Investment, exports, and economic growth in Malaysia, Thailand, and Vietnam: Descriptive and Comparative Analysis
Foreign direct investment, exports, and economic growth in Vietnam
The 'Doi Moi' reform initiated in 1986 has positioned Vietnam as a leading example of transitional economies, marking a significant turning point in its economic and social development Today, Vietnam boasts one of the fastest-growing economies globally, as noted by Ishii (2007).
Between 1994 and 2010, the average annual growth rate was approximately 7.4%, with a nominal GDP of US$ 106,427 million and a GDP per capita of US$ 1,224.3 in 2010.
Vietnam experienced significant impacts from the 1997-1998 financial and economic crises, with GDP growth dropping to 5.8% in 1998, a decline of 3.6% from 1996 In 1999, GDP growth further decreased by about 1% compared to the previous year However, the Vietnamese economy rebounded rapidly, achieving a GDP growth rate of 7.1% by 2002.
The main drivers of the economic growth were household consumption, investment and total exports in the period of 1989 - 2010 (Table 3.4 )
Table 3.4: 11te contribution l~{'compouents of' total demand to economic xrowlh iu Vietnam (percenlat.:e)
Notes: C 8 , CP, I, E, M are government consumption, household consumption, investment, exports and imports, respectively
Source: Calculated from World Development Indicators, World Bank (2012)
In 2007, fixed capital formation experienced a significant increase of 24.2%, with the fixed capital investment to GDP ratio reaching 38.3%, maintaining over 30.0% for the previous decade (World Bank, 2012) This investment growth was fueled by rising domestic credit, decreasing inflation, and lower interest rates Notably, consumption expenditure emerged as the largest contributor to GDP growth, with household consumption accounting for approximately 71% of GDP from 1989 to 2010, while exports contributed over 51% during the same period.
THble 3.5: Correlation between the ~:rowth rafei; l~/'the real GDP and its components in Vietnam, 1989-2()1() c Ir E M
Notes: C, lg, lp, Ir, E, M are the growth rates of consumption, government investment, public investment, foreign investment, exports and imports, respectively
Source: Calculated from Government Statistics Office, 20 12
Table 3.5 indicates that government investment is closely correlated with economic growth, while foreign private investment exhibits a negative relationship.
GDP growth in 1990-1997, however, for the period of 1998-2010, foreign, private investment and economic growth have a strong positive correlation Inversely, for government investment and GDP growth have a negative correlation in 1998-2010
Exports have also positively relation to economic growth in the period of 1998-
In the initial years following the 'Doi Moi' reform in 2010, government investment significantly contributed to economic growth However, as time progressed, this investment began to negatively impact economic development, leading to an increased dependence on foreign investment and exports Ultimately, during this period, Vietnam's national economy primarily relied on external forces for its growth.
Table 3.6: G'DP growth (g=A}/}), Rate r~f investment (Ill), ;:1}//, /COR in Vietnam
Source: Calculated from Government Statistics Office, 2012
Investment plays a crucial role in driving economic growth by enhancing capital stock and increasing production capacity The impact of investment on economic growth is influenced by both the investment rate (I/GDP) and investment efficiency (i1GDP/I).
Table 3.6 indicates that economic growth is primarily driven by the investment rate rather than the efficiency of investments, suggesting that the quantity of investors is more crucial than the quality of their investments The investment rate has surged to over 42%, approaching saturation levels Furthermore, the increasing ICOR indicators during the study period imply a significant decline in investment efficiency, indicating a worsening trend.
Figun' 3.8: G~DP, exports and investmellf in Viellutm, 1989-2010 (VND billion)
Figure 3.8 also indicates that, the correlation between real GDP, exports and investment is relative clearly
Table 3.7 illustrates the significant transformation of the Vietnamese economy, shifting from a traditional agricultural base to a focus on industry and services This decline in the agricultural sector is attributed to falling agricultural prices and unfavorable terms of trade in international markets.
Table 3.7: Structural change iu Vietnam /989-21JJ(J (percentage)
Source: Calculated from World Development Indicators, 2012
The demand for food grows at a slower pace than that of other goods and services, while the industrial sector's contribution to GDP continues to rise significantly.
41% in 2010 The contribution of services sector to economic growth is significant and it is concentrated on tourism
Following the 'Doi Moi' reform, Vietnamese exports experienced significant growth, reaching approximately 30% in 1991, with an average annual increase of about 19% from 1990 to 2005 However, in 2006, export growth sharply declined to -10.1%, primarily due to export quotas, trade barriers, and anti-dumping duties affecting seafood and leather products.
Figure 3.9: The J.:rowth t~fexports in Vietnam, 1990-2()/(J (percenlaJ.:I!)
In addition, exports increased slowly by the management of immanent shortcomings of economy
Joining in WTO in 2007 had helped exports back grew and reached to 11.3%
(2007) Nevertheless, Vietnamese exports also have suffered the strongest negative impact and the export of growth is decline from 11.29% in 2007 to 5.05% in 2008 by the global crisis (World Bank, 2012)
Table 3.8: E.\]JOrfs and Imports in Vietnam (perccnlaJ.:e of' GDP)
Source: Government Statistics Office, 2012 tot nghiep do wn load thyj uyi pl aluan van full moi nhat z z vbhtj mk gmail.com Luan van retey thac si cdeg jg hg
The decline in export revenues has significantly impacted exporting companies, particularly those in the manufacturing sector such as garments, footwear, furniture, and seafood Many agricultural product exporters are also facing severe challenges, putting them at risk of closure.
In the post-reform period, Vietnam consistently experiences a trade deficit, with exports falling short of imports (see Table 3.8) This trade imbalance can be attributed to several factors: high domestic consumption of imported goods, low added value in exports, the presence of import barriers imposed by other countries on Vietnamese products, and the overall low competitiveness of Vietnam's goods in both domestic and international markets.
Vietnam's economy remains primarily agricultural and reliant on natural resources, with major exports consisting of agricultural products, raw materials, and low-value processed goods due to the use of unskilled labor While the export value index experienced slow growth from 1989 to 2001, it has seen a significant increase in recent years.
Table 3.9: Export mlue iudex iu Vietnam, 1989-2010
Year 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 Export 13.5 16.6 14.4 17.9 20.7 28.1 37.7 50.3 63.3 64.8 79.9 Year 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Export 100.0 104.0 115.6 139.5 183.3 224.6 275.7 336.1 433.9 395.2 496.0 Source: World Development Indicators, 2012
Vietnam's export economy has historically relied on imported raw materials and intermediate goods, often at high prices In 1997, the country's primary exports were minerals and food; however, there was a notable shift as the export of these commodities declined and manufactured goods began to rise By 2009, manufactured products accounted for over 60% of total exports, although this sector remains largely in its infancy in Vietnam.
Figunã 3.10: Export struclure in Vietnam
97 2000 2003 2007 manufactures • mineral 11 food agricultural • others
Source: Calculated from World Development Indicators, 2012
Research Methodology
Data sources
This thesis examines the causal relationship between foreign direct investment (FDI), exports, and economic growth in Malaysia, Thailand, and Vietnam from 1989 to 2010 Data for this study is sourced from the World Bank's World Development Indicators, with a focus on annual time series and panel data due to the availability of Vietnamese indicators only after 1989 The analysis includes key variables such as real gross domestic product (GDP), real exports of goods and services (EX), and FDI.
Variable measured in constant 2000 US dollars.
Model Specification
The study employs the Vector Autoregression (VAR) model to analyze the causal relationship between foreign direct investment (FDI), exports, and economic growth, represented as U(VAR) = (GDP, EX, FDI) Prior to analysis, the variables of real exports, real FDI, and real GDP are transformed into logarithmic form as lnFDI, lnEX, and lnGDP This model aims to estimate the long-run relationship among these three variables across three countries in the Southeast Asia region.
Y = f(FDI, EX) (4.1) tot nghiep do wn load thyj uyi pl aluan van full moi nhat z z vbhtj mk gmail.com Luan van retey thac si cdeg jg hg
The model is expressed in linear form as in the following equation: lnGDPit =Po+ P1i lnFDiit + P2i lnEXit + Eit where: i is i th country and t is the time period for each country;
GDP is gross domestic product;
EX is total export values of goods and services;
FDI is foreign direct investment inflows;
P 1 and P 2 represents the long run elasticity of economic growth with respect to foreign direct investment and exports respectively; and
Po is the constant term and Et is the random error term
4.3 Estimation techniques 4.3.1 Individual economy's Granger causality test
This paper aims to investigate the causal relationship among foreign direct investment (FDI), exports, and economic growth in Southeast Asia To determine the direction of causality between these variables, Granger causality tests are employed Before conducting these tests, the study first assesses the stationarity of each time series using the Augmented Dickey-Fuller (ADF) and Phillips and Perron (PP) tests.
The Johansen cointegration test is employed to analyze the cointegration relationship among the variables By assessing the time series characteristics for each country, a Vector Autoregression (VAR) or Vector Error Correction Model (VECM) is estimated to conduct the Granger causality test Detailed presentations of each test follow.
4.3.1.1 Unit root tests for stationary time series
Preliminary analysis of the data using the Augmented Dickey-Fuller (ADF) (Dickey and Fuller, 1981) and the Phillips and Perron (PP) (Phillips and Perron,
Unit root tests are essential for determining the existence of unit roots in time series data, which can be in the level, first difference, or second difference Identifying the stationary properties of these series is crucial, as non-stationary data can render forecasting and regression analysis ineffective, leading to spurious regression results (Gujarati, 2011) Generally, a variable is considered stationary after differencing \(d\) times, indicating it is integrated of order \(d\), denoted as I(d) Most economic variables are typically integrated of order one (Asteriou and Hall, 2007).
Augmented Dickey Fuller test bases on rejecting a null hypothesis (H 0 : 8=0) of unit root (the series are non-stationary) in favor of the alternative hypothesis (Ha:
The article discusses the testing for the existence of a unit root in the time series \(Y_1\), which can represent either \(\ln GDP_1\), \(\ln EX_1\), or \(\ln FDI_1\) The optimal lag length is determined using the Akaike Information Criterion (AIC) or the Schwarz Information Criterion (SIC) A rejection of the null hypothesis of a unit root occurs when the absolute value of the Augmented Dickey-Fuller (ADF) statistic exceeds the absolute value of the critical value, indicating that the series is stationary The tests consider both random walks with and without drift, as well as time trends for each series, with three possible forms of the ADF test represented by specific equations.
~Yt = 8Yt-I + 'Lf=1 /3i~Yt-i + et pure random walk (4.3)
~Yt =a+ 8Yt-I + 'Lf=1 /3i~Yt-i + e 1 random walk with intercept (4.4)
~Yt =a+ yT + 8Yt-I +'Lf= 1 /3i~Yt-i + e 1 random walk with intercept and time trend (4.5) where:
~ is the first difference operator; a is a constant;
The article discusses a linear time trend represented by T and highlights the presence of a random error term denoted as e Additionally, it mentions the availability of the latest full download for a thesis, emphasizing the importance of accessing updated academic resources.
The Phillip-Perron (1988) test is an enhancement of the ADF test, addressing its limitations by allowing for correlated error terms and variable variance Unlike the ADF test, which assumes uncorrelated errors, the PP test adjusts the t statistic of the coefficient from the AR(1) regression to account for serial correlation in the error term The underlying regression for the Phillip-Perron test follows the AR(1) process.
Similarly, the null hypothesis of PP test is rejected as well as ADF test With bandwidth are selected by Newey-West Bartlett kernel
Most economic time series exhibit non-stationarity, necessitating tests to determine if all variables are integrated of the same order before proceeding with cointegration analysis This thesis employs the Johansen approach to investigate the cointegration relationship among exports, foreign direct investment inflows, and economic growth within a trivariate model for three Southeast Asian countries The variables lnGDPt, lnEXt, and lnFDit are considered endogenous, forming the vector Zt = [lnGDPt, lnEXt, lnFDit], which is integrated of order one, I(1) In cases with only two variables (n=2), this represents the simplest scenario for the Johansen cointegration test, while models with more than two variables can yield n-1 cointegrating vectors.
This study determines the optimal lag length by minimizing the Akaike Information Criterion (AIC) and Schwarz's Bayesian Criterion (SBC) The common approach involves estimating a Vector Autoregression (VAR) model with a large number of lags and then gradually reducing the lags by re-estimating the model until reaching zero lags Consequently, a VAR model is utilized to analyze the cointegration relationship among the variables.
Zt = AIZt-1 + A2Zt-2+ + AkZt-k + Et
It can be reformulated in a Vector Error Correction Model (VECM) as:
L1Zt = riLlZt-1 + r 2LlZt-2 + + r k-ILlZt-(k-1) + nzt-1 + Et where:
(4.8) ri=(I-AI-A2-ããã -Ak)(i= 1,2, ,k-l)andTI=-(I-AI-A2-ããã -Ak)
~ is the first difference operator;
Zt is the set of variables discussed above;
Et ~ niid (O,L), J.l is a drift parameter;
TI is a (r x r) matrix with rank r (r :S 3), r determine the number of cointegrating vector; and n = aw where a will be the speed of adjustment to equilibrium coefficients while
W will be the long run matrix of coefficients
Johansen (1988) and Johansen and Jeselius (1990) introduced two statistical tests for identifying the number of cointegration vectors The first test, known as the trace statistic (Atrace), utilizes a likelihood ratio test focused on the trace of the matrix The second test, referred to as the maximal eigenvalue statistic (Amax), is based on the maximum eigenvalue.
Atrace(r) =- T If=r+lln(l - Ar+l) Amax(r, r + 1) =- T ln (1-Ar+l) where:
X is the estimated eigenvalue; and
T is the number of observations
If the test statistic exceeds the critical value, the null hypothesis of no cointegration is rejected, supporting the hypothesis of full rank.
Cointegration among I(1) variables indicates a causal relationship, though it does not specify the direction of causality As noted by Mehrara (2007), when variables are cointegrated, the Vector Error Correction Model (VECM) is the most comprehensive method for testing Granger causality The VECM effectively addresses the limitations of VAR-based models by differentiating between long-run and short-run relationships among variables Therefore, the VECM is utilized to examine the causal relationship between the variables lnEXt, lnFDit, and lnGDPt, which can be expressed through a specified VAR model.
L1lnEXt = ~o + L~=l ~1il'1lnEXt-i + L~=l ~2il'1lnGDPt-i + L~=l ~3iL1lnFDit-i + A-2ECT2,t-I + E2t (4.12) L1lnFDit =Yo+ L~=l y 1 il'1lnFDit-i + L~=l y 2 il1lnEXt-i + L~=l y 3 iL1lnGDPt-i + A3ECT3,t-I + E3t (4.13) where:
~ is the first difference operator;
The residual Eit is assumed to be normally distributed and white noise;
ECTt-I is the one period lagged error correction term derives from the long run cointegration equation (this term will be excluded if the variables are not cointegrated); and
/ represent the deviation of the dependent variables from the long run equilibrium
If cointegration exists between two variables, error correction models will be employed to test Granger causality as outlined in equations (4.11), (4.12), and (4.13) Conversely, if the series are not cointegrated, the standard Granger causality test will utilize a vector autoregression (VAR) in first differences, omitting the error correction terms from the aforementioned equations.
The coefficients AJ, A2, and A3 in the error correction terms indicate the adjustment speed of ~lnGDPt, ~lnEXt, and ~lnFDit towards long-run equilibrium, while the lagged terms ~lnGDPt-i, ~lnEXt-i, and ~lnFDit-i reflect the short-run dynamics among these variables The Vector Error Correction Model (VECM) investigates the Granger causality relationship between exports and economic growth in both short and long runs Short-run relationships are assessed using the F Wald test for weak Granger causality, while long-run causality is evaluated through the error correction term coefficients based on t statistics Strong Granger causality is identified when short-run and long-run causal relationships align within the model, necessitating the testing of specific null hypotheses.
Short run Granger causality (F statistic) (a) H0 : L~=l a2 i = 0 or exports do not cause economic growth (b) Ho: L~=l ~zi = 0 or economic growth does not cause exports Long run Granger causality (t statistic)
(c) H0: A 1 = 0 or Granger non causality in the long run (d) H 0 : A 2 = 0 or Granger non causality in the long run Strong Granger causality
(e) H 0 : A 1 = L~=l a 2 i = 0 or exports do not strongly cause economic growth (f) H 0 : A 2 = L~=l ~zi = 0 or economic growth does not strongly cause exports
4.3.2 Panel data Granger causality test
This research employs panel data analysis to mitigate potential distortions related to size that may arise in time series analysis due to a limited number of observations The advantages of using panel data include a larger dataset, increased degrees of freedom, reduced collinearity among variables, and enhanced efficiency (Baltagi, 2001; Erdil and Yetkiner, 2006) The study investigates the causal relationship between foreign direct investment (FDI), exports, and economic growth through modern econometric techniques, which involve three key steps The first step tests for panel unit roots to determine the order of integration for each variable If unit roots are present, the second step conducts a panel cointegration test to identify any long-run relationships among the variables If a long-run relationship is established, the final step estimates a Vector Error Correction Model (VECM) to explore the Granger causal relationship between economic growth and exports, while a Vector Autoregression (VAR) model is utilized to analyze the causal links among these panel variables.
It is important to know the stationary properties of the panel variables