Mối quan hệ nhân quả giữa FDI, độ mở thương mại và tăng trưởng kinh tế ở Việt Nam bằng phương pháp ARDL kiểm định đường bao

INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020
ICYREB 2020

CAUSALITY INTERACTIONS BETWEEN FDI, TRADE OPENNESS
AND ECONOMIC GROWTH IN VIETNAM. EVIDENCE
FROM ARDL BOUNDS TESTS
MỐI QUAN HỆ NHÂN QUẢ GIỮA FDI, ĐỘ MỞ THƯƠNG MẠI
VÀ TĂNG TRƯỞNG KINH TẾ Ở VIỆT NAM
BẰNG PHƯƠNG PHÁP ARDL KIỂM ĐỊNH ĐƯỜNG BAO

Nguyen Hai Yen, MA
University of Economics, Hue University


Abstract

This paper examines the dynamic relationship between FDI, trade openness and economic
growth in Vietnam during the period of 1990-2017. This paper tests two fundamental questions
including whether FDI and trade openness indeed enhance economic growth and whether FDIled growth model is valid for Vietnam. Unlike existing studies in Vietnam which demonstrated
some limitations in terms of methodologies and presented ambiguous results in this field, this
paper applies the appropriate cointegration methodology to investigate the long-run relationship
between the variables using the autoregressive distributed lags (ARDL) bounds tests approach.
The paper then can indicate the directional causality through test for Granger dynamic causality
in the short-run. The findings of this empirical study are that there exists a long- run relationship
from economic growth to FDI, trade openness, physical capital and labour force. In the longrun, FDI and labour force have a significant and positive impact on economic growth. The shortrun dynamics further show that the Vietnamese economy convergency from a shock is relatively
quickly. The paper supports the FDI-led growth hypothesis and that Vietnam has directly benefited
from foreign trade investment inflows. The results of the Granger causality test show that there
is bi-directional causality from FDI to economic growth and trade openness to economic growth.
The paper, however, fails to confirm the direction of causality from FDI to labour force and trade
openness to labour force as the widespread belief that FDI generates significant and positive
productivity externalities for a host country. The insights of this empirical study strengthen the

case for further focus on sustainable development in policy-making.
Key words: ARDL bounds tests, economic growth, FDI, trade openness, Vietnam
Tóm tắt

Nghiên cứu về mối quan hệ nhân quả giữa FDI, độ mở thương mại và tăng trưởng kinh tế
ở Việt Nam trong giai đoạn 1990-2017 được phân tích bằng cách áp dụng phương pháp phan
phối trễ tự hồi quy (Autoregressive Distributed Lag ARDL) và phuong pháp kiểm định đuờng
bao (bounds tests). Hai câu hỏi được đặt ra là có hay khơng FDI và độ mở thương mại thực sự
thúc đẩy tăng trưởng kinh tế Việt Nam và giả thuyết mơ hình tăng trưởng dựa vào FDI có đúng
với trường hợp ở Việt Nam hay không. Phương pháp đồng liên kết được sử dụng để phân tích
245


INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020
ICYREB 2020

mối quan hệ trong dài hạn và kiểm định Granger được dùng để phân tích mối quan hệ nhân quả
trong ngắn hạn giữa các biến trong mơ hình. Kết quả nghiên cứu chỉ ra rằng, trong dài hạn, có
mối quan hệ giữa FDI, độ mở thương mại, vốn là lao động đối với tăng trưởng kinh tế. Nghiên
cứu chỉ ra được tác động tích cực và có ý nghĩa về mặt thống kê của FDI và lao động đối với
tăng trưởng kinh tế của Việt Nam trong dài hạn. Nghiên cứu ủng hộ giả thuyết tăng trưởng dựa
vào FDI đối với Việt Nam và dòng vốn FDI đã những tác động tích cực cho tăng trưởng kinh tế
của Việt Nam.
Từ khóa: ARDL kiểm định đường bao, tăng trưởng kinh tế, FDI, độ mở thương mại,
Việt Nam
1. Introduction

Trade openness and foreign direct investment (FDI) have been well known as vital factors
in stimulating the process of economic growth. Trade openness is defined as a ratio of exports
and imports to gross domestic product (GDP). Increasing imports of final goods tend to promote

competition in a local market by forcing domestic companies to adopt technological innovation
and enhance product quality to compete with imported products. For exporting companies, competition in international marketplaces encourages businesses to upgrade their competitive capabilities via innovation and adoption of advanced technologies. Therefore, an increase in trade
openness positively and significantly affects economic growth process via two mechanisms, including enhancing technology transfer and increasing competition in the domestic market.

FDI plays an important role and has a positive impact on economic growth. It is determined
to boost economic growth via the two mechanisms, including spurring competition and technology transfer. Firstly, the presence of multinational corporations as a form of FDI supports economic growth by promoting competition in the domestic market. The foreign firms have
technological advantages while domestic firms have comparative advantages by using existing
resources, local branches must effectively use the resources and adopt the new technology (De
Mello, 1997, 1999; Wang & Blomström, 1992). FDI brings the transfer of technology and
knowledge to a host country. The positive spillovers of FDI on a domestic country occur in the
form of the mobility of well-trained labour force. FDI can positively affect on economic growth
by the enhancement of efficiency.

Over the last four decades, Vietnam has witnessed rapid economic growth accompanied
by significant structural change and attainment of country status from a low- to middle-income
country. Despite having some advantages such as the young population, low-cost labour and rapid
economic growth, Vietnam continues struggling to overcome the middle-income trap (OECD,
2016). Spence (2011) found that many middle-income countries have remained in this status for
long periods and World Bank (2012) estimated that only 13 out of over one hundred middle-income countries have overcome the status and become high-income economies since 1961. Therefore, this challenge raises a priority target for sustainable growth. Among the crucial engines that
led to the Vietnam economic success, FDI and trade openness would be essential factors that
contribute to overcoming the middle-income trap and ensure sustainable development.
246


INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020
ICYREB 2020

Despite numerous empirical studies analyzing the long-run and short-run relationship between trade openness, foreign direct investment and economic growth in emerging market
economies and developing countries, few studies focus on specific analysis in Vietnam. Most
prior empirical studies in Vietnam focused on the effect of either FDI or trade openness on economic growth (Pham, 2008; Pham & Nguyen, 2013). All the empirical studies came to a similar

conclusion that FDI positively impacts on Vietnam’s economic growth. However, the previous
studies have not provided the causality direction between the variables in Vietnam (Pham, 2008;
Vu, 2008) and suffered distinct drawbacks (Pham & Nguyen, 2013; Trinh & Nguyen, 2015). The
first is that despite a large volume of empirical research in analysis between FDI, trade openness
and economic growth, there are not enough studies on the fundamental questions of causality relationship between them and determinants of economic growth in Vietnam. The second is that
these papers employed either the cointegration method developed by Engle and Granger (1987)
or the maximum likelihood test developed by Johansen (1988) and Johansen and Juselius (1990).
According to Odhiambo (2009), the techniques, however, are not good enough with a small sample size, and the appropriate method, in this case, should be bounds tests for the cointegration
approach based on Pesaran, Shin, and Smith (2001).

The current study contributes to the existing empirical literature with research findings
based on applying the cointegration methodology to examine the relationship between FDI, trade
openness and economic growth to enhance or complement those in earlier studies. Moreover, the
current study provides new insights into the non-existence causality from FDI to labour force.
That is why Vietnam should encourage FDI and strengthens the quality of labour force in a coordinated way to achieve middle-income trap through an increase in productivity and economic
growth. In addition, the study is applying to Vietnam that represents an interesting case study for
other middle-income countries, which implement incentive policies to attract FDI and reach highincome status.

The study examines the dynamic linkage between trade openness, FDI and economic
growth in Vietnam by employing effective methodology ARDL-bounds tests to investigate cointegration. Trade openness is expressed as a ratio of imports and exports to GDP, while real GDP
per capita is used as a proxy of economic growth. Labour force and physical capital are control
variables in this study. The Granger causality test is applied to test the causality direction within
the Vector Error Correction Model (VECM).

The content of the paper is organized as follows: Section 2 presents the theoretical and empirical literature review on the relationship between foreign direct investment, trade openness,
and economic growth. Section 3 provides data used in the paper. Section 4 is a methodology with
a model specification that will be employed in the paper. Section 5 describes a discussion of empirical results. Sections 6 concludes the paper.

2. Literature review

The relationship between FDI and economic growth was developed by two economic theories, namely the neoclassical theory and the new growth theory. From the neoclassical theory
perspective, FDI as driven technological progress only affects the long-run growth of the output
level with an increasing amount capital and income per capita, not on the growth rate as dimin247


INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020
ICYREB 2020

ishing returns law for the marginal product limit the growth rate in the long run (Solow, 1956,
1957). The neoclassical model then implied that the influence of FDI on the long-run growth rate
is only through exogenous variables considered as technological progress or growth rate of labour
force.

Although the neoclassical theories emphasized the effect of technological progress on economic growth, the determinants of technological progress were left as a black box. The challenges
of the new growth theories framework were to assess the determinants of technological progress.
Therefore, the new growth theories developed the contribution of FDI on economic growth
through the analysis in the crucial engine of growth including innovation creation. The primary
consideration of FDI’s influence on economic growth is due to the high volume of physical capital
and the level of technology. However, the main channel to contribute to economic growth of FDI
is a rising level of technology rather than an accumulation of physical capital in a host country
(Borensztein, De Gregorio, & Lee, 1998). Moreover, the transfer of technology in the form of
FDI is another mechanism to stimulate economic growth due to technology diffusion in the existence of other sources such as accumulation of human capital, knowledge transfers and modern
knowledge.

In fact, a function of technology is the main difference between the neoclassical theory and
new growth theories. The former considered technology progress as an exogenous variable, the
latter explained the role of FDI on growth as the investment spill-over coming from a variety of
sources such as human capital, development and research.

On the other hand, international trade was not considered as an engine of economic growth

during the twentieth century due to a dominance of protectionist theories in this period, at least
during the time of Adam Smith. The theoretical framework of the neoclassical analysis fails to
provide the impact of trade openness, measured as the ratio of net exports to GDP, on the growth
of economic progress. The linkage between trade and economic growth, however, has remained
an important subject of debate in research and policy implications.

The positive effect of trade openness on economic growth was supported by the new growth
theories (Lucas, 1988; Romer, 1986). Trade openness enhances economic growth based on two
mechanisms as technological progress and a rise in competition in a local market. Firstly, technological change is a key driver of improvements in the way to access international markets and
gain efficiencies from economies of scale through a higher level of trade openness. It is more
likely that the economy is growing faster with the adoption of more efficient techniques (Barro
& Sala-i-Martin, 1995; Romer, 1986). Another way trade openness positively spreads across a
domestic market is to promote competition for local businesses. Under an open trade policy, domestic branches should be pushed to be more innovative and creative to compete with imported
products in the local market. The advocates of the export-led growth point out that exports are
one of the determinants of economic growth because local firms face competition from foreign
firms. The local branches, therefore, enhance efficiency via learning by doing or emulation of
international competitors (Ben-David & Loewy, 1998; Bonelli, 1992).

The empirical framework analysing the influence of FDI and trade openness on economic
growth has been a subject of many economists. However, the interrelationship between these
248


INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020
ICYREB 2020

variables is still debated with contrasting results based on different methodologies and different
data sets.

By using two stage least squared regression structural analysis in seven Mediterranean

countries including Algeria, Egypt, Jordan, Morocco, Syria, Tunisia and Turkey over the period
1982- 2009, Marc (2011) found that human capital and exports are the main factors to stimulate
economic growth in these countries but FDI inflows have a significantly negative effect on economic growth. Applying bounds testing ARDL approach to cointegraiton, Belloumi (2014) investigated the dynamic causality between FDI, trade and economic growth in Tunisia from 1970
to 2008. The author found that although FDI and trade promote economic growth in the long run,
there is a lack of significant Granger causality from FDI and trade to economic and from economic
growth to FDI and trade in the short run. In line with the cointegration test of ARDL bounds tests,
Maha and Zghidi (2017) indicated the long run relationship between FDI, international trade and
economic growth for 15 selected Middle Eastern and North African countries. The authors, however, failed to confirm the granger cause of FDI to economic growth and provided the important
role of positive spill-over externality of FDI. By applying an ARDL model for cointegration,
Alalaya (2010) studied FDI, trade openness and economic growth for Jordan over the period of
1990-2008. The author found that there was unidirectional causal effect between trade and FDI
to economic growth.

Rahman (2009) applied the cointegration technique to test the influence of FDI, exports,
remittances on real GDP of selected Asian countries including Bangladesh, India, Pakistan and
Sri Lanka over the period of 1976-2006. The author found the cointegration relationship among
the variables of these countries.

Using a new endogenous growth model and Granger causality test among trade openness
and economic growth in India, Hye and Lau (2015) found that trade openness negatively impacts
on economic growth while capital formation and human capital positively effect on real GDP in
the long-run. In the short run, trade openness has a positive sign. The Granger causality test confirmed both hypotheses including openness-led growth and human capital-led growth in short
and long run.

The positive relationship between FDI and real GDP also was confirmed in some case
studies in Asia such as in Thailand (Roy & Mandal, 2012), Pakistan (Rahman & Shahbaz, 2013),
Indonesia (Suyanto, Bloch, & Salim, 2012).

Regarding Vietnam, much work has been done on the analysis of either relationship between FDI and economic growth or trade and economic growth. Indeed, the results are inconclusive and have been under debate. Varamini and Vu (2007) examined the linkage between FDI
and economic growth of Vietnam during 1989-2005 by applying the ordinary least square method,

and the authors found the positive impact of FDI inflows on Vietnam’s economic growth. Vu
(2008) employed the impact of FDI inflows on Vietnam’s economic growth through the seven
economic sectors and the results showed positive effects of FDI inflows on economic growth
and labour productivity; however, the distribution of FDI influences across seven sectors is different. By investigating the relationship between FDI, exports and economic growth in Vietnam
using annual time series data from 1986 to 2015 based on ARDL and error correction model,
249


INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020
ICYREB 2020

Nguyen (2017) demonstrated the positive and negative impact of FDI and exports on economic
growth in the long run respectively. Pham and Nguyen (2013) examined the linkage between
FDI and economic growth from 1991 to 2012 and found that there existed a long run relationship
of bidirectional causality between the two variables. Pham (2008) examined the investment-led
growth and export-led growth hypotheses by using two VAR models from 1986 to 2007. The author found investment to be the main driver of Vietnam economic growth while the influence of
exports on economic growth is very small.

Sothan (2016) found that both FDI and exports-led growth in Vietnam economics by applying panel cointegration for 21 Asian countries. While Trinh and Nguyen (2015) found the positive impact of FDI and trade openness on Vietnam economic growth based on cointegration
methodology over the period from 1990 to 2013.

Based on the analysis of the literature above, it is no doubt our study will fill the gap and
contribute to the empirical framework by investigating the relationship between FDI, trade openness, economic growth, physical capital and labour in Vietnam with the ARDL bounds tests
methodology.

3. Data sources and description of variables

In this study, we employed time series data on GDP, FDI, trade openness, physical capital,
and labour force from 1990 to 2017. These data come from different sources, including the World
Bank Indicators (WBI) and Penn World Table 9.1. The data of labour, FDI, import, export, and

physical capital are sourced from WBI and the rest is obtained from the Penn World Table 9.1.

Economic growth is the rise of the productions and services of a country that is measured
by real GDP at a constant 2010 US dollar. The variable real GDP per capita is designated as Y
while the value of real gross foreign direct investment inflows, denoted as FDI, is measured as
net inflows (BoP, current US$). Trade openness (TO) is captured by the ratio of net exports to
GDP. Physical capital is measured as gross capital formation and it is expressed in constant 2010
US dollar. Labour denoted by L is measured the volume of labour force. The descriptive statistics
are given in Table 1. For all five distributions, the Jarque-Bera statistics do not allow to formally
reject the hypothesis of normality and show that the time series data is normally distributed.
Variables

Mean

Median

Maxium

Minimum
Std.dev.

Skewness
Kurtosis

Jarque-Bera
Probability

Observation

Table 1. Descriptive Statistics

lnY

12.334
12.320
13.233

11.252
0.614

lnFDI

lnTO

21.722

lnK

lnL

23.739

17.625

11.63

21.64

17.31

12.681

21.457

12.88

23.369

13.562

1.152

0.571

19.008

23.615
24.8

0.905

17.61

17.848
0.169

-0.102

-0.345

-0.452

-0.665

-0.212

1.864

0.935

2.1

2.356

1.915

28

28

28

28

1.753
0.394

2.43

2.009

0.627

0.35

250

2.5

0.308

1.791
0.384
28


INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020
ICYREB 2020

4. Econometric Methodology

This paper is to empirically test the dynamic causal relationship between FDI, trade openness and economic growth of Vietnam with two stages. Firstly, by implementing the efficient
ARDL bounds test approach of cointegration, we examine the existence of the long-run relationship between these variables. Secondly, the Granger causality test is applied to investigate the
short-run causality direction within the vector error correction model (VECM).

The paper also considers the role of capital formation and labour force in the economic
growth model as other determinants of economic growth to avoid omitted variables.

A general model of economic growth, trade openness and foreign direct investment can
be presented as follows:

Yt = f(FDIt,TOt,Kt,Lt)

The equation can be expressed as follows:

(1)

lnYt = α+β1lnFDIt + β1lnTOt + β3 lnKt + β4lnLt + μt

(2)

Where lnY, lnFDI, lnTO, lnK and lnL stand for the natural log of economic growth, the
natural log of trade openness, the natural log of foreign direct investment, the natural log of
capital information and the natural log of labour force respectively. The variables all are taken
into natural logs due to two reasons. First, based on the natural logs of all variables, the coefficients of the cointegration models can be interpreted as elasticities. Second, with the natural
logs of variables, the first difference of all variables can be interpreted as growth rates.

4.1. Unit root test

The first step is to test whether the variables are stationary or non-stationary to examine
the long-run relationship and conduct causality tests between these variables. Three tests of unit
root including Augmented Dickey-Fuller (ADF) by Said and Dickey (1984), Phillips- Perron
(PP) unit root test by Perron and Phillips (1988) and the Dickey-Fuller Generalised Least Squares
(DF-GLS) by Elliott, Rothenberg, and Stock (1996) are used in this study. The drawback of ADF
and PP unit root tests is that it has low power if the variables are stationary but with a root close
to the non-stationary boundary. Compared to the above unit root tests, DF-GLS unit root test
gains power in the presence of GLS detrended data and size, DF-GLS then has the best performance (Elliott et al., 1996).

4.2. ARDL bounds test for cointegration

This study empirically analyzes the long-run relationship and the short-run dynamic causality among the variables (economic growth, FDI, trade openness, physical capital and labour force)

based on the autoregressive distributed lag (ARDL) technique. The model of ARDL was developed by Pesaran and Shin (1998) and Pesaran et al. (2001).

The ARDL model is chosen to conduct an empirical analysis about the long-run relationship
and dynamic causality interactions between the variables because the ARDL model has three advantages in comparison with the previous cointegration procedure of Johansen (1988) and Johansen and Juselius (1990). Firstly, the ARDL model has a flexible order of integration that means
the same order integration of the variables as the past technique is not necessary. The order integration of the variables could be I(0) or I(1). The second advantage of the ARDL model is that
251


INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020
ICYREB 2020

the model can be applied efficiently in a small sample size with some ARDL tests. Lastly, unbiased regressors can be estimated in the long-run model in the presence of ARDL methodology
(Harris & Sollis, 2003).
Based on the ARDL approach, the models of this study are as follows:

252


INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020
ICYREB 2020

Where all variables are defined previously and formed as natural log, D is the first difference and is the error terms. The coefficients such as , are the long-run multipliers while the
short-run multipliers are presented in the coefficients of difference terms. The approach of the
ARDL bounds methodology has some steps. Firstly, based on the ordinary least square (OLS)
method, the above five equations ((3) - (7)) are to estimate to test the long- run relationship between the variables. The hypothesis testing of the presence of the long-run relationship is a
procedure in the bounds test based on a joint F test with the null hypothesis of no cointegration
against the alternative hypothesis. Critical values are given in two sets as follows: the first
level is an upper critical bound value based on the assumption of the ARDL model that is integrated of order one while the second one is a lower critical bound value with the assumptions
of integration of order zero of the ARDL model. The null hypothesis of no long-run relationship is rejected if the value of F test statistics exceeds the upper critical bounds value, while Ho
cannot be rejected when the calculated F falls below the lower bounds values. Otherwise, the

test is inconclusive.

4.3. Granger causality

Once the long-run cointegration relationship existed, the short-run and long-run Granger
causality among the variables is the next step. The main idea of Granger causality test is that a
time series Xt Granger-causes another time series Yt if the information of time series Xt can be
used for the prediction of future values of time series Yt better than the already existing information in the past of Yt series (Friston, Moran, & Seth, 2013).

The Granger causality test has been applied in this paper due to the advantages of this
method over other alternative methods.
The long-run conditional ARDL model can be presented as the following:

Where all the variables are defined previously. The last step of ARDL model is to estimate
the VECM model to investigate the short-run relationship in the presence of the long-run relationship found previously.

253


INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020
ICYREB 2020

Where is the coefficient of the error correction term (ECT) and it represents the convergence speed of the model into the equilibrium.

5. Findings of the study

5.1. Unit root tests

To conduct the causality and long-run test with time series analysis, a stationary test for all
variables is essential. The order of cointegration of the variables must be identified based on the

unit root tests to satisfy the assumption of the ARDL bounds test that the variables are cointegrated
at the order I(0) or I(1). In fact, if the variable is I(2), the F-test would be spurious because two
critical values of bounds tests are computed by Pesaran et al. (2001) with variable assumptions
of I(0) or I(1). The stationary results are shown in Table 2. This study applies three methods of
unit roots tests, including ADF, the Phillips-Peron test and AF-GLS test. The results provide that
all variables are non-stationary at the level and become stationary in their first difference.
Table 2: Unit roots tests on level and first difference of log levels of variables

Variables
lnY

ADF

DFGLS

-2.69

-2.32

-1.5003

-2.44

-2.44

-2.487

lnK

-2.48

lnL

PP

-1.5

lnFDI
lnTO

Levels

-2.45

-2.4364
-2.45

ADF

First difference
PP

DF-GLS

Integration

-1.66

-4.88***

-4.87***

-5.02***

I(1)

-2.21

-3.6**

-3.66**

-3.29**

I(1)

-1.57
-0.69

-3.74**
-3.81**
-3.3*

-3.72**
-4**

-3.3*

-3.65**
-3.7***

-3.46**

Note: (***), (**), and (*) denotes 1%, 5% and 10% level of statistical significance

I(1)
I(1)
I(1)

5.2. Cointegration

In this section, ARDL bounds test is a procedure used to examine the long run relationship between the variables. The first step, the maximum order of lags on the first difference in
Eq.(3)-(7) is chosen by using Akaike Information Criterion (AIC) and Schwartz Bayesian
Criterion (SBC)

The analysis for ARDL bound testing is illustrated in Table 3. The bounds tests show that
F-statistics of ARDL test are higher than upper critical bounds. The F statistics are significant at
the 1% level as economic growth, FDI, trade openness, physical capital and labour are dependent
variables respectively. This means that the null hypothesis of no cointegration among the variables
in Eq.(3)-(7) is rejected. The results support the existence of long run relationship between variables over the period of 1990-2017.

254


INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020
ICYREB 2020

Table 3: Results from bounds tests
Dependent variable
AIC lags
F-statistic

FlnY (lnY/ lnFDI,lnTO,lnK,lnL)
Significant level
1%
5%

1

5.09***

Critical values

Decision

Cointegration

Lower bounds I(0) Upper bounds I(1)
3.74
2.86

5.06
4.01

Note: (***), (**), and (*) denotes 1%, 5% and 10% level of statistical significance, respectively. The optimal lag is determined by AIC. Lower and upper-bounds critical values are obtained from Pesaran et al. (2001)

The maximum lag order is chosen for two in the ARDL model based on the Akaike information criteria (AIC). The consideration of each variable as dependent variables and the estimated
F-statistics values were represented in Table 3 that presents the long run relationship among the
variable based on the ARDL technique.

From the results in Table 3 , there is no long run relationship when GDP is a dependent

variable, the calculated F- statistics is less than the lower critical bounds value at the 5% level of
significance. The rest of variables, on the other hand, are cointegrated as capital, trade openness
and foreign direct investment are considered as dependent variables due to higher value of F statistics than upper bounds at the 5% level. This implies that there is long run relationship
amongst the variables.

Once a long run relationship as cointegration is established, equation (8) is estimated using
the following the ARDL (1,0,0,0,0) specification for GDP, FDI, TO, K and L respectively as the
order of the ARDL model selected by using AIC (see Table 4).

The results indicate that the estimated coefficients of the long-run relationship are significant for foreign direct investment and labour. Foreign direct investment positively impacts economic growth in the long run. A 1% increase in foreign direct investment boosts economic growth
by 0.076% keeping others constant. Similarly, labour force has a positive (3.42) and significant
impact on economic growth. Considering the effect of trade openness and physical capital, the
two variables are insignificant at the 5% level. The model failed to indicate any long run relationship between degree of trade opennes, physical capital and economic growth. The degree of
trade openness does not stimulate economic growth.

255


INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020
ICYREB 2020

Table 4: Estimated Long Run Coefficients using the ARDL Approach- ARDL(1,0,0,0,0)
C

Variable

Ln(FDI)
Ln(TO)

Coefficient

t-Statistic

0.076***

3.016

-48.06***
-0.103

Ln(K)

-1.44

-0.063

Ln(L)

-1.11

3.42***

R-squared

4.24

0.998

F-statistic

DW-statistic

-4.23

-

2433.05

-

1.344

Probability
0.0004
0.0066
0.1638
0.279

0.0004
-

0.0

-

-

Having estimated the long-run model, the next step is to determine the short-run dynamic
parameters by estimating error correction model associated with the long- run ARDL estimates.
The results of short-run dynamic are obtained from Eq.(9) and given in Table 5.

In the short run, the impact of trade openness on economic growth is significant but negative. Foreign direct investment positively and significantly effects on economic growth.

The estimate of the lagged error correction term is significant at the 5% level and negative
as expected, which confirms the long-run relationship between the variables. The significance
and negative sign of the estimated error-correction term reveals the speed of adjustment from the
short-run towards the long-run equilibrium. The speed of adjustment towards long run equilibrium
is high with its values at -0.775 showing that any short-run shock in economic growth in the previous year converges back to the long-run equilibrium path by approximately 77.5% in the current
year. In the long run, FDI, trade openness, capital formation and labour force significantly impact
economic growth in the Granger sense. The finding implies that the dynamic causality runs
through the error correction term from FDI, trade openness, capital and labour force to economic
growth.

C

Variable

D(Ln(FDI))
D(Ln(TO))
D(Ln(K))
D(Ln(L))
ECT(-1)

Table 5: Short Run Results
Coefficient
0.014

0.094***
-0.22**
-0.054

2.84**

-0.775**

R-squared

0.628

DW-statistic

1.769

F-statistic

5.354

256

t-Statistic
0.57
3.61

-2.27
-0.65
2.46

0.0222
-

Probability
0.5763
0.0019
0.039

0.5247
0.0236
-

0.0022
-


INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020
ICYREB 2020

The stability of the long run coefficient is essential and detected by the short-run dynamics.
Though based on estimated model of ECM in Eq(9), the cumulative sum of recursive residuals
(CUSUM) and the CUSUM of square (CUSUMSQ) tests are obtained to detect the stability of
parameters (Pesaran & Pesaran, 1999). The results for CUSUM and CUSUMSQ tests are illustrated in Figure 1 and Figure 2. The results indicate the presence of stable coefficients because
graphs of CUSUM and CUSUMSQ statistic are within critical bounds at the 95% confidence interval of parameter stability.
Figure 1: Plot of Cumulative Sum of Rescursive Residuals

Figure 2: Plot of Cumulative Sum of Squares of Recursive Residuals

257


INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020
ICYREB 2020

The diagnostic tests can be obtained from Table 6. The problem of non-normality of residual
term does not exist in the results. The results show no problem of autoregressive conditional heteroskadasticity and the serial correlation was not found. Besides, from the results of residual diagnostic, there is no evidence of white heteroskedasticity.
statistic

Table 6: Results of diagnostic tests

Breusch-Godfrey serial correlation test

χ2 statistic

Probability

4.75

0.31

1.035

White Heteroskedsticity test
Jarque-Bera test

0.399

3.5

0.176

The outcomes of the short- run Granger causality tests are given in Table 7. Based on the
F-statistics of the explanatory variables, the short-run causality results show bidirectional causality

between FDI and economic growth, between trade openness and economic growth, between physical capital and FDI.

In short-run, there is bi-directional Granger causality between trade openness and economic
growth. That means the positive and significant impact of trade liberalization on growth. Indeed,
the effect is not only contributed from degree of trade, but also justified from the capital movement in the type of FDI due to the existence Granger causality from FDI to trade openness.

These seems to confirms the existence of trade openness and economic growth nexus. The
hypothesis of FDI-led growth also is confirmed in the finding. This means that local businesses
got benefit from multinational corporations through spillover effect mechanism.

The unidirectional Granger causality runs from FDI and labour to trade openness and from
economic growth and trade openness to physical capital.

It is interesting that there is no significant Granger causality from FDI and trade openness
to labour force.
Table 7: Granger short run and long run causality tests

Dependent
varibale

D(lnY)

D(lnY)

—-

D(lnTO)

11.23***

D(lnFDI)

4.69**

D(lnFDI)

F statistics
D(lnTO)

D(lnK)

D(lnL)

Direction
of causality

6.61**

6.22**

0.009

4.66**

FDI”Y; TO”Y; L”Y

5.06**

——-

0.024

3.64**

Y”TO; FDI”TO;
L”TO

——-

0.73

11.8***

0.97

D(lnK)

7.99***

14.9***

2.72*

——-

0.93

D(lnL)

2.09

0.81

0.518

0.22

———

258

Y”FDI; K”FDI

Y”K; FDI”K;
TO”K


INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020
ICYREB 2020

The non-existence of causality from FDI to labour force in the empirical study fail to confirm the widespread belief that FDI can generate positive productivity externalities for a host
country. Instead of generating positive spillovers in case of the mobility of well-trained labour
channel, foreign-owned firms skim trained labour force provided by domestic firms and free ride
the good labour force provided by domestic firms at least in the short run.

6. Conclusion

This study examines the causal dynamic relationship between FDI, trade openness, physical
capital, labour and economic growth in Vietnam for a period of 1990-2017. This paper focuses
on a fundament question whether FDI and trade openness spurs economic growth and whether

FDI-led growth hypothesis is supported by time series data in Vietnam. The paper employs ARDL
bounds tests to analyze the existence of the long-run interaction between the above variables and
the Granger causality within VECM to test causality direction among the variables. Evidence of
cointegration in the paper implies that the long-run relationship between the variables as economic
growth is the dependent variable.

Overall, the empirical results of this study show that FDI and trade openness spur economic
growth in Vietnam.

REFERENCES

Alalaya, M. (2010). ARDL models applied for Jordan trade, FDI and GDP series. European
Journal of Social Sciences, 13(4), 605-616.
Hill.

Barro, R. J., & Sala-i-Martin, X. (1995). Economic growth New York: New York : McGraw

Belloumi, M. (2014). The relationship between trade, FDI and economic growth in Tunisia:
An application of the autoregressive distributed lag model. Economic Systems, 38(2), 269-287.

Ben-David, D., & Loewy, M. (1998). Free Trade, Growth, and Convergence. Journal of
Economic Growth, 3(2), 143-170.
Bonelli, R. (1992). Growth and productivity in Brazilian industries: Impacts of trade orientation. Journal of development economics, 39(1), 85-109.

Borensztein, E., De Gregorio, J., & Lee, J. W. (1998). How does foreign direct investment
affect economic growth? Journal of International Economics, 45(1), 115-135.

De Mello, L. R. (1997). Foreign direct investment in developing countries and growth: A
selective
survey.

The
Journal
of
Development
Studies,
34(1),
1-34.
doi:10.1080/00220389708422501

De Mello, L. R. (1999). Foreign direct investment-led growth: evidence from time series
and panel data. Oxford Economic Papers, 51(1), 133-151.

Elliott, G., Rothenberg, T., & Stock, J. (1996). Efficient tests for an autoregressive unit
root. Econometrica, 64(4), 813.
259


INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020
ICYREB 2020

Engle, R. F., & Granger, C. W. J. (1987). Co-Integration and Error Correction: Representation, Estimation, and Testing. Econometrica, 55(2), 251-276.

Friston, K., Moran, R., & Seth, A. (2013). Analysing connectivity with Granger causality
and dynamic causal modelling. Current Opinion in Neurobiology, 23(2), 172-178.
Harris, R., & Sollis, R. (2003). Applied time series modelling and forecasting.

Hye, Q. M. A., & Lau, W.-Y. (2015). Trade openness and economic growth: empirical evidence from India. Journal of Business Economics and Management, 16(1), 188-205.

Johansen, S. (1988). Statistical analysis of cointegration vectors. Journal of economic dynamics and control, 12(3), 231-254.

Johansen, S., & Juselius, K. (1990). Maximum Likelihood Estimation and Inference on
Cointegration - With Applications to the Demand for Money. Oxford Bulletin of Economics and
Statistics, 52(2), 169.

Lucas, R. E. (1988). On the Mechanics of Economic Development. Journal of Monetary
Economics, 22(1), 3-42.

Maha, K., & Zghidi, N. (2017). Foreign Direct Investment, Trade, and Economic Growth
in MENA Countries: Empirical Analysis Using ARDL Bounds Testing Approach. Journal of the
Knowledge Economy, 10(1), 397-421.

Marc, A. (2011). Is foreign direct investment a cure for economic growth in developing
countries? Structural model estimation applied to the case of the south shore Mediterranean countries. Journal of International Business and Economics, 11(4), 32-51.

Nguyen, N. T. K. (2017). The Long Run and Short Run Impacts of Foreign Direct Investment and Export on Economic Growth of Vietnam. Asian Economic and Financial Review, 7(5),
519-527.

Odhiambo, N. M. (2009). Energy consumption and economic growth nexus in Tanzania:
An
ARDL
bounds
testing
approach.
Energy
Policy,
37(2),
617-622.
doi:10.1016/j.enpol.2008.09.077

OECD. (2016). Economic Outlook for Southeast Asia, China and India 2016: Enhancing

Regional Ties: OECD Publishing.

Perron, P., & Phillips, P. C. B. (1988). Testing for a Unit Root in Time Series Regression.
Biometrika, 75(2), 335-346.

Pesaran, B., & Pesaran, H. (1999). Microfit 4.1 interactive econometric analysis. In: Oxford
University Press, Oxford.

Pesaran, M. H., & Shin, Y. (1998). An autoregressive distributed-lag modelling approach
to cointegration analysis. Econometric Society Monographs, 31, 371-413.

Pesaran, M. H., Shin, Y., & Smith, R. J. (2001). Bounds testing approaches to the analysis
of level relationships. Journal of Applied Econometrics, 16(3), 289-326.

Pham, A. M. (2008). Can Vietnam’s Economic Growth be Explained by Investment or Export: A VAR Analysis. Paper presented at the VDF Working Paper Series.
260


INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020
ICYREB 2020

Pham, T. A., & Nguyen, T. H. D. (2013). Bound Testing Approach to Cointegration: A Reexamination of FDI and Growth in Vietnam. Journal of Economic Development(218), 94-112.

Rahman, M. (2009). Contributions Of Exports, Fdi And Expatriates’remittances To Real
Gdp Of Bangladesh, India, Pakistan And Sri Lanka. Southwestern Economic Review, 36, 141153.

Rahman, M. M., & Shahbaz, M. (2013). Do Imports and Foreign Capital Inflows Lead
Economic Growth? Cointegration and Causality Analysis in Pakistan. South Asia Economic Journal, 14(1), 59-81.

Romer, P. M. (1986). Increasing Returns and Long-Run Growth. Journal of Political Economy, 94(5), 1002-1037.

Roy, S., & Mandal, K. (2012). Foreign Direct Investment and Economic Growth: An Analysis for Selected Asian Countries. Journal of Business Studies Quarterly, 4(1), 15-24.

Said, S. E., & Dickey, D. A. (1984). Testing for unit roots in autoregressive-moving average
models of unknown order. Biometrika, 71(3), 599-607.

Solow, R. (1956). A contribution to the theory of economic growth. The quarterly journal
of economics, 70(1), 65-94. doi:10.2307/1884513

Solow, R. (1957). Technical Change and the Aggregate Production Function. The Review
of Economics and Statistics, 39, 312.

Sothan, S. (2016). Foreign direct investment, exports, and long-run economic growth in
Asia panel cointegration and causality analysis. International journal of economics and finance,
8(1), 26-37.
Spence, M. (2011). The Next Convergence. The Future of Economic Growth in a Multispeed
World. New York: Farrar, Straus and Giroux.

Suyanto, Bloch, H., & Salim, R. A. (2012). Foreign Direct Investment Spillovers and Productivity Growth in Indonesian Garment and Electronics Manufacturing. The Journal of Development Studies, 48(10), 1397-1411.
Trinh, N. H., & Nguyen, Q. A. M. (2015). The Impact of Foreign Direct Investment on
Economic Growth: Evidence from Vietnam. Developing Country Studies, 5(20), 1-9.

Varamini, H., & Vu, A. (2007). Foreign direct investment in Vietnam and its impact on
economic growth. International journal of business research, 7(6), 132-139.

Vu, T. B. (2008). Foreign direct investment and endogenous growth in Vietnam. Applied
Economics, 40(9), 1165-1173.

Wang, J.-Y., & Blomström, M. (1992). Foreign investment and technology transfer: A simple model. European Economic Review, 36(1), 137-155. doi:10.1016/0014-2921(92)90021-N

World Bank. (2012). 2012 World Development Report on Jobs. In. World Bank, Washington, DC.

261