JRSSEM 2022, Vol. 01, No. 7, 842 – 852
E-ISSN: 2807 - 6311, P-ISSN: 2807 - 6494
DOI : 10.36418/jrssem.v1i7.103
BILATERAL TRADE ANALYSIS OF ASEAN AND CHINA
COUNTRIES IN ACFTA COOPERATION (GRAVITY MODEL
APPROACH)
Devi Tri Wulandari1*
Lilis Yuliati2
Siti Komariyah3
1,2,3University of Jember
e-mail: devitwulandari@gmail.com1, lilisyuliati.feb@unej.ac.id2, Sitikomariy[email protected].id3
*Correspondence: devitwulandar[email protected]
Submitted: 27 January 2022, Revised: 07 February 2022, Accepted: 18 February 2022
Abstract. At the end of 2001 ASEAN and China agreed on free trade in Bandar Sri Begawan, Brunei
Darussalam, known as the ASEAN-China Free Trade Agreement (ACFTA). Periodically, ASEAN and
China make agreements, one of the goals of which is to eliminate or cut barriers to trade in goods,
both tariffs and non-tariffs. Under ACFTA, tariff reduction began in July 2005 and aims to cut import
duties to zero by 2010 on about four thousand types of goods for the relatively developed ASEAN
countries namely Thailand, Malaysia, Singapore, Indonesia, the Philippines and Brunei. The Gravity
Model predicts trade based on distances and interactions between countries in terms of their
economic size. The Gravity Model in economics imitates Newton's law of gravity which also takes
into account the physical distance and size between two objects. The application of this model to
explain economic phenomena regarding the interaction between the two countries has been
widely carried out by economists. The study uses panel data from China and ASEAN6 in the 2010-
2020 research period with ASEAN6 exports to China as the dependent variable, and the
independent variables include the GDP of the destination country and the country of origin,
economic distance proxied in the form of transportation tariffs, exchange rates and economic
openness. Panel data regression analysis was used to see the effect of the independent variable on
the dependent variable by determining the best model (common effect, fixed effect, random effect)
and the classical assumption test performed was the multicollinearity test and the
heteroscedasticity test. The results showed that the GDP of destination and origin countries,
distance, and exchange rates significantly affected the export value of ASEAN6 to China.
Meanwhile, economic openness has no significant effect on the value of ASEAN6 exports to China.
Keywords: gravity model; international trade; ASEAN; ACFTA; panel data regression.
Devi Tri Wulandari, Lilis Yuliati, Siti Komariyah | 843
DOI : 10.36418/jrssem.v1i7.103
INTRODUCTION
The economic growth of a country
cannot be separated from the role of
international trade which is one of the
factors that can be used as a driving force
for economic growth or an increase in the
value of GDP. To improve trade relations
with countries in ASEAN, at the end of 2001
ASEAN agreed on free trade with China
within the framework of ACFTA (ASEAN-
China Free Trade Agreement) in Bandar Sri
Begawan, Brunei Darussalam and fully
implemented in 2010. Exports of ASEAN
countries China, based on data from
ASEAN Statistics, from 2010 to 2019 was in
the top rank compared to other countries.
This can be seen in Figure 1, where China
occupies the top chart and shows a positive
trend since 2010. In addition to China's
exports which have been ranked the
highest in the last ten years, the value of
China's imports to ASEAN also ranks at the
top (Zhang, Yang, Wang, Zhan, & Bian,
2020). This can be seen in Figure 2 which
shows a positive trend of Chinese imports
into the ASEAN market.
Figure 1 shows that the exports of
ASEAN countries had the highest number
of exports to China during 2010 to 2019.
Likewise with the number of imports to the
ASEAN market (Webb, Strutt, Gibson, &
Walmsley, 2020), China was in the top rank
compared to other countries during 2010
to 2019. Of course the increase in export
value and Chinese imports in the ASEAN
market occur in line with the policy of
implementing ACFTA cooperation (Chen &
Lombaerde, 2019), one of which is that
there is no tariff for imports. In addition, the
ease of trade between China and ASEAN
countries is also influenced by the distance
between countries which is not far when
compared to the distance between non-
ASEAN countries. Since January 1, 2010,
China and ASEAN-5 plus Brunei have
removed tariffs on 7000 product categories
covering 90% of traded goods (Li et al.,
2016).
Figure 1. Exports of goods from ASEAN countries to countries in 2010-2019
(Source: ASEAN Statistical Yearbook 2020)
844 | Bilateral Trade Analysis of ASEAN and China Countries in ACFTA Cooperation (Gravity
Model Approach)
Figure 2. Imports of goods from countries to ASEAN markets in 2010-2019
(Source: ASEAN Statistical Yearbook 2020)
International trade is created because
There are differences in production from
one country to another. Smith argued that
trade between two countries was based on
absolute advantage. Smith in his theory
believes that all countries will benefit from
free trade which causes the world's
resources to be used efficiently and
maximize welfare. But in Smith's view there
is a paradox that most countries impose
many restrictions on the free flow of
international trade. While in reality, trade
restrictions are only recommended by
some industries and trade unions who feel
threatened by imported products (Howse
& Langille, 2012) Classical international
trade theory has received criticism from
modern theory because classical theory
cannot explain why there are differences in
the production function between two
countries. Modern trade theory from the
Hecker-Ohlin (HO) model explains that
countries export what they are most
efficient and produce the most (Espinoza,
2020). Economic integration in general is
the removal (removal of) economic barriers
between two or more economies
(countries). Operationally, discrimination is
defined as deprivation and political unity
(policy) such as norms, rules, procedures.
These instruments include import duties,
taxes, currencies, laws, institutions,
standardization, and economic policies.
There are two approaches that are
particularly useful as literature in
international trade policy, namely the
Gravity Model which predicts trade based
on the distance between countries and the
interaction between countries in terms of
their economic size, and the Computable
General Equilibrium Models (Babatunde,
Begum, & Said, 2017). The Gravity Model in
economics imitates Newton's law of gravity
which also takes into account the physical
distance and size between two objects. The
application of this model to explain
economic phenomena regarding the
interaction between the two countries has
been widely carried out by economists. In
practice, export activities are economic
Devi Tri Wulandari, Lilis Yuliati, Siti Komariyah | 845
activities that are directly related to other
countries. (Mulyadi, Zhang, Dutzer, Liu, &
Deng, 2017) have conducted a similar
study, which in his research found that the
GDP of export destination countries had a
significant effect and had a positive sign. In
line with the theory described by the
Gravity Model where the GDP of the
destination country increases, exports to
that country will increase. In addition,
research also conducted by (Abbas &
Waheed, 2015) in Pakistan found that
distance has a negative relationship to
Pakistan's exports. However, (Naudé,
Bosker, & Matthee, 2010) found that the
economic distance variable had a positive
and significant effect on new coal exports.
Referring to the exposure of empirical
studies that have been carried out
previously, this study wants to further
examine the role of ACFTA cooperation,
China's trade cooperation with ASEAN6
member countries (Brunei Darussalam,
Indonesia, Malaysia, Philippines, Singapore,
and Thailand) with the Gravity Model.
METHODS
Objects and Types of Research Data The
Objects in this study are ASEAN6
member countries (Brunei Darussalam,
Malaysia, Philippines, Singapore, Thailand,
Indonesia) and China. The selection of
research objects is based on the ACFTA
agreement which is fully valid only with 6
ASEAN6 countries in 2010. The data used in
this study is panel data which includes
export data of each ASEAN6 country to
China, China's GDP and GDP of ASEAN6
countries, distance economy, exchange
rates and economic openness.
The type of data used in this research is
annual data from 2010 to 2020. The
selection of the research period is based on
the full implementation of the ACFTA
cooperation agreement. Data sources are
taken from IMF, World Bank, ASEAN
Statistics, comtrade.org and
distanceworld.com. With the number of
countries studied are 6 countries and a time
period of 11 years using annual data, then
the amount of data used in this study is 66
data.
Research Model Specifications This
research
Model specification adopts research
that has been done previously by (Agung,
Ishak, Asngari, & Bashir, 2019); (Irshad, Xin,
Shahriar, & Arshad, 2017). The variables
used as indicators in this research are
Export, GDP, Distance, Exchange Rate, and
Economic Openness. The econometric
model that will be used in this study is
written in equation 6 as follows.
𝑙𝑛𝑋𝑖𝑗𝑡= 𝛼0+ 𝑎1𝑙𝑛(𝐺𝐷𝑃𝑖𝑡. 𝐺𝐷𝑃
𝑗𝑡)
+ 𝑎2𝑙𝑛𝐷𝐼𝑆𝑇𝑖𝑗𝑡 + 𝑎3𝑙𝑛𝐸𝑋𝐶
𝑗𝑡
+ 𝑎4𝑂𝑃𝐸𝑁
𝑗𝑡 + 𝑒𝑖,𝑡
Where:
lnX : exports of each ASEAN6 country to
China in t converted into natural logarithm
(ln)
lnGDP : product of China's GDP with GDP
of ASEAN6 countries in year t converted
into natural logarithm (ln)
lnDIST : transportation costs from each
ASEAN6 country to China in year t
converted into natural logarithm (ln)
lnEXC : exchange rate of national
currencies of ASEAN6 countries per US$,
average per period converted into natural
846 | Bilateral Trade Analysis of ASEAN and China Countries in ACFTA Cooperation (Gravity
Model Approach)
logarithm (ln)
OPEN : economic openness of ASEAN6
countries
i : Indonesia, Singapore, Brunei,
Malaysia, Philippines, Thailand
j : China
t : 2010 – 2020
Data Analysis Method The
Method used in explaining the analysis
of ASEAN 6 bilateral trade with China is the
Least Square Panel (PLS). The Panel Least
Square (PLS) method will provide an
explanation related to the formulation of
the problem in this study (Hair, Sarstedt,
Ringle, & Mena, 2012). PLS is an estimation
method that uses panel data, which is a
combination of time series and cross
section data so that more data will be
observed than time series or cross section.
In addition, the use of panel data will make
the regression results tend to be better
than regressions that only use time series
or cross section. In using the Least Square
Panel (PLS), there are several approaches
used to estimate the model parameters,
namely the Common Effect, Fixed Effect
and Random Effect approaches.
The next test after getting the best
model results in this study will be a classical
assumption test which generally consists of
autocorrelation, multicollinearity, and
heteroscedasticity tests. However, in this
study, the classical assumption test that
was only used was the Multicollinearity and
Heteroscedasticity test. This is because in
panel data which is a combination of time
series and cross section there will be no
autocorrelation because autocorrelation
only occurs in time series data. In addition,
if the best model chosen later is the Fixed
Effect, then the autocorrelation test does
not need to be carried out because the
Fixed Effect does not require free equations
from autocorrelation (Nachrowi, 2006; 334).
Operational Definition of Variables
Several variables used in this study have
various units and their respective
operational definitions. Some of the
variables used also consist of various
different sources. The operational
definitions of the variables used in this
study consist of the following:
a. Xij is the export of each ASEAN6 country
to China with units of million US$. Net
export data obtained from
comtrade.org.
b. GDPij is the product of the GDP of China
and each ASEAN country6 (Irshad et al,
2018). GDP is a proxy for the size of the
economy in the gravity model. The GDP
used is real GDP based on the base year
2015. The unit used is US$ and the data
is taken from the World Bank.
c. DISTij is a proxy for transportation costs
in conducting international trade. The
value of the economic distance of the
country of origin to the country of
destination is obtained from the
calculation of the geographical distance
of the capital cities of the two countries
multiplied by the nominal GDP of the
destination country (China) in the last
research period divided by the total
nominal GDP of the destination country
(China) in year t (Liu et al., 2021). Country
distance data is taken from
distanceworld.com and GDP data is
taken from the World Bank.
Devi Tri Wulandari, Lilis Yuliati, Siti Komariyah | 847
d. EXCj is the average domestic currency
exchange rate (domestic currency) per
period in US$ units in ASEAN countries6.
The exchange rate used is the real
exchange rate. Exchange rate data is
taken from the IFS website.
e. OPENj is the percentage of economic
openness of each ASEAN country6
expressed in percent. Data obtained
from the World Bank.
RESULTS AND DISCUSSION
China's trade growth rate has increased
rapidly since 2001, when the country joined
the WTO and held two initial meetings to
discuss the establishment of the ASEAN–
China Free Trade Area (ACFTA). More
specifically, the average annual growth rate
in bilateral trade from 2001 to 2008 was
about 30%. In 2011, ASEAN became China's
third largest trading partner behind the
United States and the European Union.
China and ASEAN consider the period
between 2002 and 2009 to be a transition
period before the completion of ACFTA.
During this period, the tariffs imposed on
goods traded between China and ASEAN
will be lowered gradually. Under the goods
trade agreement, the reduction in tariffs
began in July 2005 and aims to cut import
duties to zero by 2010 on about four
thousand types of goods for six relatively
developed ASEAN countries namely
Thailand, Malaysia, Singapore, Indonesia,
the Philippines and Brunei, and to 5 % in
2015 for other ASEAN members, namely
Vietnam, Laos, Cambodia and Myanmar
(Yang & Martinez-Zarzoso, 2014).
The development of bilateral trade
between China and ASEAN6 since the
implementation of the ACFTA policy which
cut import duties to zero since 2010 can be
seen in Figure 3 China's net exports in
Figure 3 show a positive trend during the
period of zero import duty. Although at the
beginning of 2010 – 2011 China's net
exports to ASEAN6 countries decreased,
but in the following year (2011 to 2015) the
development of China's trade with ASEAN6
countries showed an increase. The
development of this positive trend does
not seem to be going well because net
exports declined again in 2015 to 2017 for
Brunei, Thailand, Malaysia and Singapore,
while Indonesia and the Philippines still
showed a positive trend. In the final year of
the study period (2019-2020), Indonesia,
Singapore and Brunei trade with China
showed a decline. Based on ASEAN Key
Figures 2020, the Covid-19 pandemic and
the movement restrictions it causes have a
significant impact on trade and supply
chains around the world, including ASEAN.
This ultimately resulted in weakening
international trade (ASEAN Secretariat,
2020). In its latest forecast, the World Trade
Organization (WTO) projects a 9.2% decline
in trade volumes by the end of 2020 (WTO,
2020).
Devi Tri Wulandari, Lilis Yuliati, Siti Komariyah | 848
DOI : 10.36418/jrssem.v1i7.103
Figure. 3 Net Exports of China – ASEAN6 2010 – 2020
(Source: comtrade.org)
The positive trend of China's bilateral
trade with ASEAN6 countries cannot be
separated from the country's economic
conditions. This can be seen from the
development of the country's GDP which in
this study can be seen in Figure 4. Based on
data obtained from the World Bank, the
GDP of ASEAN6 countries showed a
positive trend from 2009 to 2019. Similar to
trade, the Covid-19 pandemic 19 also
affects the GDP of ASEAN countries6. Based
on a report from ASEAN Key Figures 2021,
the continuous increase in GDP per capita
from 2000-2019 has decreased due to the
COVID-19 outbreak in 2020 (Tailor, 2020).
Figure. 4 GDP of ASEAN Countries6 2009 – 2020
(Source: World Bank)
Panel Regression Model Selection Test
Before entering the Panel Regression
estimation, a series of panel regression
model selection tests will be carried out
based on the significance test of the model
by comparing the Common Effect (CEM),
Fixed Effect (FEM) and Random Effect
(REM). The significance test of the model is
carried out using the Chow test for the first,
Devi Tri Wulandari, Lilis Yuliati, Siti Komariyah | 849
where this test will see the best model from
the comparison of the CEM model and
FEM. Furthermore, Hausman test will be
carried out to see the best model from the
comparison of the FEM model and REM.
The first model selection test was
conducted, namely the Chow test. The
results of the Chow test in Table 1 show a
cross-section F 0.0232 which means less
than 0.05, so it can be determined that the
best model between CEM and FEM is FEM.
Furthermore, in testing the model selection
between FEM and REM which was carried
out with the Hausman test. The results of
the Hausman test can be seen in Table 2
where the table shows that the probability
value of a random cross section is 0.0435,
which means less than 0.05. From the
Hausman test, it can be determined that
the best model between FEM and REM is
FEM.
Table 1. Test Results Chow
Effects Test
Statistics
df
Prob.
Cross-section F
5.487768
(5.56)
0.0004
Cross-section Chi-square
26.318308
5
0.0001
Source: author's preparation
Table 2 Results of the Hausman
Test Summary
Chi-Sq.statistics
Chi-Sqdf
Prob.
Cross-section random
11.042514
4
0.0261
Source: author's preparation
From the model testing that has been
carried out using the Chow test and
Hausman test, the best model for research
using panel data regression is Random
Effect (REM) which will be discussed in the
next subsection.
Panel Data Regression Test with
Random Effect
After the researcher selects the best
model, then estimation is done by panel
data regression test using the Random
Effect. The results of the panel data
regression estimation using the Random
Effects show the results where there are 3
independent variables having a probability
value of less than 0.05, which means that
these three variables significantly affect the
exports of each ASEAN6 country to China
with an alpha of 5%, while one independent
variable is economic openness does not
have a significant effect on exports of each
ASEAN6 country6 to China. The results of
the panel data regression estimation with
the Random Effect can be seen in Table 3
as follows.
Table 3. Regression Panel Data with Random Effect Model
Coefficient
Std. Error
t-Statistic
Prob.
-66.23974
4.982195
-13.29529
0.0000
1.663232
0.092921
17.89936
0.0000
-0.327274
0.154613
-2.116729
0.0384
-0.288815
0.048801
-5.918280
0.0000
850 | Bilateral Trade Analysis of ASEAN and China Countries in ACFTA Cooperation (Gravity
Model Approach)
-0.002
.
_
_
(logx) with an alpha of 5%, namely
loggdp, logdist and logexc which are
indicated by probability values less than
0.05 alpha. The loggdp variable has a
positive effect on imports of 1.663232
which means that when the GDP of China
and ASEAN6 countries has increased by 1%,
the exports of each ASEAN6 country to
China have increased by 1.67%. The second
variable that has an effect on logx is logdist,
which in this study is the distance between
two countries as a proxy for transportation
costs. Logdist has a significant negative
effect on exports of each ASEAN6 country
to China. The effect of logdist is 0.327274,
which means that if the cost of
transportation from each ASEAN6 country
to China increases by 1%, the export of
each ASEAN6 country to China decreases
by 0.33%. The next independent variable
that affects logx is logexc. This variable has
a significant negative effect on exports of
each ASEAN6 country to China by
0.288815, which means that if the nominal
exchange rate of each ASEAN6 country
increases by 1%, the export of each
ASEAN6 country to China decreases by
0.29%. The next variable is openness which
according to the estimation results has a
negative effect on exports of each ASEAN6
country to China and but has no significant
effect because the probability of this
variable is greater than 0.05.
The classical assumption test used in
this study is the multicollinearity and
heteroscedasticity test. From the
multicollinearity test, it can be concluded
that the data does not experience
multicollinearity, which means that there is
no attachment between the independent
variables used in the study. This can be seen
from the values listed in Table 4, where the
table shows that there are no values
between variables smaller than 0.8.
Table 4. Test Results Multicollinearity
LOGGDP
LOGDIST
LOGEXC
openess
LOGGDP
1
0.0500776636729
0.621123320778
0.0115028304456
LOGDIST
0.0500776636729
1
0.119401910807
0.0546518595241
LOGEXC
0.621123320778
0.119401910807
1
-0.599506469682
openess
0.0115028304456
0.0546518595241
-0.599506469682
1
CONCLUSIONS
Overall it can be concluded that the
direction of the results of this study is in line
with the Gravity Model and analysis can be
formulated in several conclusions as
follows:
a. GDP of origin and destination countries
has an effect on international trade,
especially in ASEAN6 countries and
China which have established trade
cooperation in ACFTA.
b. The distance between countries that
have trade cooperation, which in this
study is proxied in the form of
transportation costs between the two
Devi Tri Wulandari, Lilis Yuliati, Siti Komariyah | 851
DOI : 10.36418/jrssem.v1i7.103
c. countries, has an effect on determining
whether the value of international trade
is large or small.
d. The exchange rate of the domestic
currency against the US$ is still one of
the influential variables in determining
international trade.
REFERENCES
Abbas, Shujaat, & Waheed, Abdul. (2015).
Pakistan’s potential export flow: The
gravity model approach. The Journal of
Developing Areas, 367–378.
Agung, Arjun Saka, Ishak, Zulkarnain,
Asngari, Imam, & Bashir, Abdul. (2019).
The effect of ASEAN-Korea free trade
agreement (AKFTA) on Indonesia trade:
A gravity model approach. Jurnal
Ekonomi Pembangunan, 17(1), 1–7.
Babatunde, Kazeem Alasinrin, Begum,
Rawshan Ara, & Said, Fathin Faizah.
(2017). Application of computable
general equilibrium (CGE) to climate
change mitigation policy: A systematic
review. Renewable and Sustainable
Energy Reviews, 7(8), 61–71.
https://doi.org/10.1016/j.rser.2017.04.0
64
Chen, Lurong, & Lombaerde, Philippe De.
(2019). ASEAN between globalization
and regionalization. Asia Pacific
Business Review, 5(5), 729–750.
https://doi.org/10.1080/13602381.201
9.1652982
Espinoza, Arturo. (2020). Peruvian
preferential trade agreements and its
impacts on non-traditional exporting
firms: an intensive and extensive margin
approach.
Hair, Joe F., Sarstedt, Marko, Ringle,
Christian M., & Mena, Jeannette A.
(2012). An assessment of the use of
partial least squares structural equation
modeling in marketing research.
Journal of the Academy of Marketing
Science, 4(3), 414–433.
Howse, Robert, & Langille, Joanna. (2012).
Permitting pluralism: the seal products
dispute and why the WTO should
accept trade restrictions justified by
noninstrumental moral values. Yale J.
Int’l L., 37, 367.
Irshad, Muhammad Saqib, Xin, Qi, Shahriar,
Saleh, & Arshad, Hamza. (2017). A panel
data analysis of China’s trade pattern
with OPEC members: Gravity model
approach. Asian Economic and Financial
Review, 8(1), 103–116.
Li, Jiwei, Galley, Michel, Brockett, Chris,
Spithourakis, Georgios P., Gao,
Jianfeng, & Dolan, Bill. (2016). A
persona-based neural conversation
model. ArXiv Preprint ArXiv:1603.06155.
Liu, Yawen, Cui, Qi, Liu, Yu, Zhang, Jinzhu,
Zhou, Meifang, Ali, Tariq, Yang, Lingyu,
Feng, Kuishuang, Hubacek, Klaus, & Li,
Xinbei. (2021). Countermeasures
against economic crisis from COVID-19
pandemic in China: An analysis of
effectiveness and trade-offs. Structural
Change and Economic Dynamics, 5(9),
482–495.
https://doi.org/10.1016/j.strueco.2021.
09.017
Mulyadi, Arie, Zhang, Zhe, Dutzer, Michael,
Liu, Wei, & Deng, Yulin. (2017). Facile
approach for synthesis of doped carbon
electrocatalyst from cellulose
nanofibrils toward high-performance
metal-free oxygen reduction and
hydrogen evolution. Nano Energy, 32,
336–346.
852 | Bilateral Trade Analysis of ASEAN and China Countries in ACFTA Cooperation (Gravity
Model Approach)
https://doi.org/10.1016/j.nanoen.2016.
12.057
Naudé, Wim, Bosker, Maarten, & Matthee,
Marianne. (2010). Export specialisation
and local economic growth. World
Economy, 33(4), 552–572.
https://doi.org/10.1111/j.1467-
9701.2009.01239.x
Tailor, Nimisha. (2020). Competition in the
new ASEAN economy. Journal of
Southeast Asian Economies, 37(3), 313–
326.
Webb, Mike, Strutt, Anna, Gibson, John, &
Walmsley, Terrie. (2020). Modelling the
impact of non-tariff measures on
supply chains in ASEAN. The World
Economy, 4(8), 2172–2198.
https://doi.org/10.1111/twec.12955
Yang, Shanping, & Martinez-Zarzoso,
Inmaculada. (2014). A panel data
analysis of trade creation and trade
diversion effects: The case of ASEAN–
China Free Trade Area. China Economic
Review, 29, 138–151.
https://doi.org/10.1016/j.chieco.2014.0
4.002
Zhang, Linfeng, Yang, Hangjun, Wang, Kun,
Zhan, Yi, & Bian, Lei. (2020). Measuring
imported case risk of COVID-19 from
inbound international flights---A case
study on China. Journal of Air Transport
Management, 8(9), 101–118.
https://doi.org/10.1016/j.jairtraman.20
20.101918
© 2022 by the authors. Submitted
for possible open access publication
under the terms and conditions of the Creative
Commons Attribution (CC BY SA) license
(https://creativecommons.org/licenses/by-sa/4.0/).