The impact of free trade agreement on trade flow of goods in vietnam

UNIVERSITY OF ECONOMICS

INSTITUTE OF SOCIAL STUDIES

HO CHI MINH CITY

THE HAGUE

VIETNAM

THE NETHERLANDS

VIETNAM - NETHERLANDS
PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

THE IMPACT OF FREE TRADE AGREEMENT
ON TRADE FLOW OF GOODS IN VIETNAM

BY
NGUYEN QUANG HUY

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

HO CHI MINH CITY, DECEMBER 2014


UNIVERSITY OF ECONOMICS

INSTITUTE OF SOCIAL STUDIES

HO CHI MINH CITY

THE HAGUE

VIETNAM

THE NETHERLANDS

VIETNAM - NETHERLANDS
PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

THE IMPACT OF FREE TRADE AGREEMENT
ON TRADE FLOW OF GOODS IN VIETNAM
A thesis submitted in partial fulfilment of the requirements for the degree of
MASTER OF ARTS IN DEVELOPMENT ECONOMICS

By
NGUYEN QUANG HUY

Academic Supervisor:
Assc. Prof. Dr. NGUYEN TRONG HOAI

HO CHI MINH CITY, DECEMBER 2014


ACKNOWLEGDEMENT
I would like to acknowledge and send heartfelt gratitude to Assc. Prof. Dr Nguyen
Trong Hoai who is my academic supervisor. His advice and recommendation have helped
me to overcome my insufficient knowledge in trade theories and econometric techniques.
Therefore, I can continue our work and finish this study.
This thesis cannot be completed without the fine guidance from the VNP Scientific

Committee. I would like to thank sincerely to Dr Truong Dang Thuy.
I would like to address my deep gratitude to Dr Pham Khanh Nam, Dr Nguyen
Minh Duc, Dr Pham Thi Thu Tra, and Dr Dinh Cong Khai for their supporting comments.
I would like to give a great thank to Ms Xuan Hong, Mr Tam, Mr Huy, and Mr Quy
in supporting and providing information, lab room for my writing and preparation.
On my writing process, I received great support from Ms Yen, Mr Anh, and Mr
Chinh, my classmates at class 19. Without their support and sharing, it is really difficult to
cope with the huge pressure in finishing thesis on time.


ABSTRACT
This study devotes to analyze the effect of free trade agreement on
Vietnam’s trade flow of goods by establishing gravity model for 184 trading
partners between 1990 and 2012. Basing on the theoretical foundation and previous
empirical papers, the FTA is expected to have positive relationship with trade flow
of member countries. In details, two countries being in a same FTA will trade much
more than those without in a same FTA. The results from the current study also find
out that FTA’s estimated coefficients are consistently positive.
Relating to the methodology, two main problems discussed in gravity model
are endogeneity of FTA and zero-value in trade. Firstly, the thesis points out that
FTA is exogenous variable in the gravity model. Secondly, sample selection model,
Poisson Pseudo Maximum Likelihood, and Fixed-effect model are used to solve the
zero problems in model. The empirical results from all estimation models are
consistent with each other in term of sign of FTA’s coefficient.
For policy implication, the thesis proposes that FTA is a good trade policy
for Vietnam because it can help to improve the total bilateral trade value among
FTA members. However, FTA does not impact on trade flow only, but it also has
effect on other aspects of Vietnam economy such as wage structural, investment.
Those issues are beyond the objective of this study.

Key words: Free Trade Agreement, gravity model, total bilateral trade


TABLE OF CONTENTS
CHAPTER 1

INTRODUCTION ..................................................................... 1

1.1 Problem Statement .............................................................................. 1
1.2 Research Objectives ............................................................................ 4
1.3 Research questions .............................................................................. 4
CHAPTER 2

LITERATURE REVIEW ............................................................... 5

2.1 Theoretical literature ........................................................................... 5
2.1.1 Trade Theories ............................................................................... 5
2.1.2 Free Trade Agreement .................................................................... 8
2.1.3 The role of free trade agreement ..................................................... 9
2.2 EMPIRICAL LITERATURE ............................................................ 16
2.2.1 Trade Flow ................................................................................... 16
2.2.2 Free Trade Agreement .................................................................. 18
2.2.3 Control ling factors ...................................................................... 19
2.2.4 Conceptual Framework ................................................................ 21
2.2.5 Empirical study on estimation model............................................ 24
CHAPTER 3 OVERVIEW OF VIETNAM’S FREE TRADE AGREEMENT ......... 26
3.1 ASEAN Free Trade Agreement (AFTA)............................................ 26
3.2 ASEAN-People’s Republic of China Comprehensive Economic
Cooperation Partnership (ACFTA) ....................................................................... 28
3.3 ASEAN- Australia and New Zealand Free Trade Agreement

(AANZFTA) ......................................................................................................... 31
3.4 ASEAN-India Free Trade Agreement (AIFTA) ................................. 33
3.5 ASEAN-Republic of Korea Comprehensive Economic Cooperation
Agreement (AKFTA) ............................................................................................ 34
3.6 Japan-Vietnam Economic Partnership Agreement (VJEPA) .............. 37
CHAPTER 4

RESEARCH METHODOLOGY ................................................. 39

4.1 Model Specification and Estimation Method ..................................... 39
4.1.1 Model for solving endogenous problem ....................................... 41
4.1.2 Model for Solving Heteroscedasticity and Zero-Trade Value ....... 44


4.1.3 Log-form gravity model: .............................................................. 45
4.1.4 Multiplicative-form gravity model ............................................... 49
4.2 Variable Definition ............................................................................ 51
4.2.1 Dependent Variable in Model Specification ................................. 51
4.2.2 Independent variables in the Model Specification......................... 51
4.3 Data Collection .................................................................................. 56
4.3.1 Sources of data ............................................................................. 56
4.3.2 Sample Selection .......................................................................... 56
CHAPTER 5

EMPIRICAL RESULT ................................................................. 60

5.1 Testing Results .................................................................................. 60
5.1.1 Endogenous testing for FTA......................................................... 60
5.1.2 Testing for Multi-collinearity ....................................................... 60
5.2 Empirical Results .............................................................................. 60

5.2.1 Log-form gravity model ............................................................... 60
5.2.2 Multiplicative-Form Gravity Model ............................................. 66
CHAPTER 6

CONCLUSION ............................................................................. 70

6.1 Conclusion ........................................................................................ 70
6.2 Policy Implication ............................................................................. 72
6.3 Limitation .......................................................................................... 72
REFERENCE

........................................................................................................ 74

APPENDICE

80


LIST OF TABLES
Table 3-1:ACFTA's Tariff Elimination schedule for Vietnam ........................................ 29
Table 3-2: Preferential Tariff Rate commitment by Vietnam for AKFTA ...................... 35
Table 4-1: Model Specifications Summary .................................................................... 50
Table 4-2: Variable Definition ....................................................................................... 58
Table 4-3: Variables Descriptive Statistics ..................................................................... 59
Table 5-1: Testing endogeneity of FTA ......................................................................... 60
Table 5-2: Regression Result for Fixed-Effect Model .................................................... 62
Table 5-3: Regression Result for Sample Selection Model ............................................. 65
Table 5-4: Testing Results for Collinearity Problem in sample selection model ............. 66
Table 5-5: Testing results for Interaction Terms in Sample Selection Model .................. 66
Table 5-6: Regression Results for PPML model Dependent variable Tvjt ...................... 67

LIST OF FIGURES
Figure 1-1: Trade share of FTA countries and non-FTA countries .............................................. 1
Figure 1-2: Trade share among Vietnam FTAs ........................................................................... 2
Figure 1-3: Trade Share of FTA countries, EU, and USA ........................................................... 2
Figure 2-1: The Relationship Between FTA and Trade Flow .................................................... 13
Figure 2-2: Conceptual Framework .......................................................................................... 23
Figure 3-1: Trade Pattern between Vietnam and ASEAN Members, 1990-2012 ....................... 26
Figure 3-2: Change in Trade Flow Vietnam and ASEAN Countries, Excluding Brunei ............. 27
Figure 3-3: Trade Pattern between Vietnam and ACFTA members, 1990-2012 ........................ 30
Figure 3-4: Change in Trade Flow between Vietnam and ACFTA Members ............................. 31
Figure 3-5: Trade between Vietnam and Australia-New Zealand, 1990-2012............................ 32
Figure 3-6: Change In Trade Flow Between Vietnam And AANZFTA Members ..................... 33
Figure 3-7: Trade between Vietnam and India, 1990-2012 ........................................................ 33
Figure 3-8: Change in Trade between Vietnam and India, 1990-2012 ....................................... 34
Figure 3-9: Trade between Vietnam-Korea, 1990-2012 ............................................................ 36
Figure 3-10: Change In Trade Flow Between Vietnam And AKFTA Members......................... 36
Figure 3-11: Trade Between Vietnam and Japan, 1990-2012 .................................................... 37
Figure 3-12: Change in Trade Flow Between Vietnam and Japan ............................................. 38
Figure 4-1: Total Bilateral Trade between Vietnam and US ...................................................... 43
Figure 4-2: ASEAN Trade Balance with China......................................................................... 44


ABBREVATION
ACFTA

ASEAN Free Trade Agreement

AFTA

ASEAN Free Trade Agreement

AJFTA

ASEAN-Japan Free Trade Agreement

AIFTA

ASEAN-India Free Trade Agreement

AKFTA

ASEAN-Korea Free Trade Agreement

ASEAN

Association of Southeast Asian Nation

BTA

Bilateral Trade Agreement

CER

Closer Economic Relation

EEC

European Economic Community

EMU

Economic and Monetary Union

FTA

Free Trade Agreement

NFTA

North America Free Trade Agreement

TPP

Trans-Pacific Partnership

VJEPA

Vietnam-Japan Economic Partnership Agreement

USA

the United States of America


CHAPTER 1

INTRODUCTION

1.1 Problem Statement

According to World Trade Organization, Vietnam is now official member of
eight free trade agreements which are signed and into force; and, Vietnam is also
launching negotiation with a number of countries and economic groups to establish
free trade agreement such as TPP, ASEAN-EU FTA. In 2012, trade balances between
Vietnam and other members of ASEAN Free Trade Agreement (AFTA), recorded at US $3.443 billion, and Vietnam – China trade balance, members of ACFTA, is – US
$16.397 billion (Customs Handbook on International Merchandise Trade Statistics of
Vietnam, 2012). In stark contrast, the net export of Vietnam to the Australia and New
Zealand, members of ASEAN-ANZ FTA, is US $ 1.2258 billion in 2012(computed
base on Customs Handbook on International Merchandise Trade Statistics of Vietnam,
2012. It can be observed that trade patterns of Vietnam among FTA partner are
different.
As can be seen from the Figure 1-1, share of FTA countries was around 50
percent of Vietnam total trade over the period of 1990-2012, thus the FTA-trading
partners play essential role in the trade of Vietnam. However, from 2000 afterwards,
the share of non-FTA countries gradually rose; this can be explained by the increase in
trade between Vietnam and European countries and the United States of America
presented in Figure 1-2

100.00%
80.00%
60.00%
40.00%
20.00%
0.00%
1990
1991
1992
1993
1994
1995

1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012

Percentage

Figure 1-1: Trade share of FTA countries and non-FTA countries

Year
Share of FTA countries

Page | 1

Non-FTA share

Source: Author’s Construction

Figure 1-2: Trade Share of FTA countries, EU, and USA
80%
70%

60%
50%
40%
30%
20%
10%
0%

Share of FTA countries

US share

EU share

Other

Source: Author’s Construction

Furthermore, among Vietnam FTAs, AFTA share was decreasing from around
30 percent in 1996 to 15 percent in 2012. The decreasing trend can be attributed from
the increase in share of AFTA plus. For instance, the trade between Vietnam and
China, member of ASEAN- China FTA, increased nearly three times from 2000 to
2012. Another positive change after signing FTA is with Korea to establish ASEANKorea FTA. Yet, the change in trade share of other AFTA plus are not significantly.
Therefore, a question posed is that whether the impact of FTA on Vietnam trade flow

is significant or not.
Figure 1-3: Trade share among Vietnam FTAs
50%
45%
40%
35%
30%
25%
20%
15%
10%
5%
0%

Non-FTA share

China Share

ASEAN share

AANZ share

India Share

Korea Share

Japan share

Source: Author’s Construction
Page | 2

Relating to academic perspective, the free trade is espoused in improving the
trade and welfare of signing countries. This belief is developed from the absolute
advantage by Adam Smith, comparative advantage by David Ricardo, HeckscherOhlin model by Eli Heckscher and Bertil Ohlin to Paul Krugman with economies of
scale and product differentiation. On the other hand, Viner (1950) not only accepted
the gain from free trade but he also pointed out the loss if countries build trading
blocs. The “trade diversion” is the terminology for the diverting trade from countries
to countries. Therefore, question should be asked is how trade flow of Vietnam is
affected by her free trade agreements.
There are many papers investigated the relationship between free trade
agreement and bilateral trade. On the one hand, Hur, Alba &Park (2009) conducted a
research to evaluate the Hub-and-Spoke problem in free trade agreements by using
panel data analysis of 96 countries from 1960 to 2000; and concluded that in spite of
the existing of overlapping free trade agreement, export also increased. Other relevant
paper by Baier & Berstrand (2009) which gives comprehensive contribution to the
empirical model to investigate the effect of free trade agreement on international trade
flow also found the positive linkage. On the other hand, Aitken (1972) does an
empirical research on trade creating and trade diverting after the establishment of EEC
and EFTA. He found that EEC increase trading value of member countries
significantly by trade creating and external trade diverting, whereas the effect of
EFTA is not considerably because of trade diverting. Besides, Bhavish, Jugurnath,
Stewart & Brooks (2007), evaluate ASEAN FTA, APEC, CER, MERCOSUR and
NFTA, assert that ASEAN FTA, CER increase significantly its members’ trade, while
MERCOSUR, NFTA and APEC created the trade diversion. As a result, the demand
for study for international trade effect on specific context needs a critical exploration.
However, papers analyses the linkage between free trade agreement and trade flow in
Vietnam is limited. For instance, Thai (2006) used gravity model to calculate the trade
between Vietnam and twenty-three European countries. The paper found the
determinants in trading such as economic characteristics, exchange rate volatility and

the demand of destination. However, Free trade Agreement is not mentioned in the
paper because Vietnam and those countries have not signed Free trade Agreement yet.
Page | 3


Other paper investigates the impact of ASEAN Free Trade Agreement (AFTA) on
Vietnam’s economy is by Fukase & Martin (2001). The authors concluded that impact
of AFTA is not significant. Agricultural sector gets benefit from that free trade
because of exporting opportunity to ASEAN market, while some other industries is
hurt and need to be protected due to ASEAN competitors. The paper applied
simulation methodology to draw the conclusion, and does not analyze the dynamic
effect of AFTA. Therefore, this thesis will depict the role of free trade agreement on
bilateral trade in goods of Vietnam by applying gravity model approach, and panel
data which developed by previous papers.
1.2 Research Objectives
The main objective of thesis is:
 To investigate the impact of free trade agreement on trade flow of goods in
Vietnam.
1.3 Research questions
To attain the research objective of this thesis, there are two primary research
questions are as follow
1. Do free trade agreements impact on trade flow of goods in Vietnam?
2. How much free trade agreements impact on trade flow of goods in Vietnam?

Page | 4


CHAPTER 2

LITERATURE REVIEW

2.1 Theoretical literature
2.1.1 Trade Theories
The first theoretical theory on international trade come from the idea of Adam
Smith in his book named “An Inquiry into the Nature and Causes of the Wealth of
Nations”, published in 1776. Adam Smith disagreed with the ideas of mercantilists
who believed that trade only benefit to the home country at the cost of trading partner.
By conceding that countries can gain from trade, Adam Smith pointed out that gain
from trade is attributed to the efficiently reallocation resource within country through
the absolute advantage of country. Each country will be more effectively and
efficiently produce certain kind of commodities than those of other countries, thus the
absolute production cost will be lower in country. Therefore, country will reallocate
its resources to specialize in producing commodities which have absolute advantage,
and exchange for product with absolute disadvantage (Salvatore, 2012). In final, both
countries can consume the greater number of commodities than those of before trade.
Besides the increase in volume of consumption, Schumacher (2012) asserted that
Adam Smith theory also attributed the gain from trade to the increasing in
international competitiveness. This is not equivalently discussed in many textbooks.
However, absolute advantage is claim for its explanation why one country whose
absolute advantage is completely unavailable can trade with other countries. That is
the problem solved by theory called comparative advantage, or Ricardian model.
Another influential theoretical foundation developed based on absolute
advantage is comparative advantage, or Ricardian model named after Ricardo, the first
economist mentioned the concept. In Ricardian model, the difference in relative labor
productivity is the primary reason for international trade between countries, because
the model assumes that labor is only one factor of production; and labor can move
freely between sectors within country. Although the model is claimed as its simplified
assumption, its contribution is considerable on explaining trade pattern of countries.
Firstly, Ricardian model pointed out that country exports a good to other country
because its amount of other good given up to produced that good in exporting country

Page | 5


is lower than that in partner country; this concept is called comparative advantage.
Comparative advantage has helped to explain why a country having no absolute
advantage can also gain from trade, in other words, the central determinant in trade is
relative productivity of goods within country in compare to that of other country. In
comparative advantage, international trade can benefit trading countries because
import activities in the model are considered similarly to indirect production
(Krugman, Obstfeld, & Melitz, 2012). Indirect production means that country can
specialize in a good, and then export that good in exchange for importing other good
at lower relative cost than domestic producing. Secondly, the comparative advantage
is an initial concept in order to develop the cost advantage of country (Krugman,
Obstfeld, & Melitz, 2012). In details, the relative low productivity country has lower
wage, and then leads to lower cost relative to cost in high productivity country. This
concept is the explanation for notion that high wage country can export goods to low
wage country because of high productivity compensation in high wage. Finally,
Ricardian model implies that all countries can benefit from trade based on its
disputing assumptions
According to Ricardian model, countries will certainly benefit from
participating in international trade; however, the counter argument is that the “one
factor assumption” is unrealistic. Therefore, the specific-factors model breaks one
factor assumption of Ricardian model by extending specific factor of production in
each sector within a country. Together with mobile labor, each sector consists of one
production factor which cannot move to other sector. The simple specific factor model
exploits capital and land as two specific factors. The specific- factor model presents
that the country can be potentially better off by trading. In details, there are winner
who gain and loser who lose from trade. The model stated that specific factor in the
exporting sector of each country gains from trade, whereas specific factor in importing

sector loses from trade; besides, mobile factor, namely labor, may gain or lose which
is depend on the consumption of labor on goods in each sector (Krugman, Obstfeld, &
Melitz, 2012). In general, Krugman, Obstfeld, & Melitz (2012) asserted that
international trade benefits countries as a whole because an increase in the wider
choice for consumption can compensate the sector hurt. In conclusion, the income
Page | 6


distribution effect helps to explain the benefit and cost of international trade, and give
reason why many countries maintain the trade barrier to protect their domestic sector
from foreign comer.
Jones & Neary (1982) asserted that specific-factor model explained the
international trade in short run, and could transform to Heckscher- Ohlin model
because specific factors are mobile in the long run. The Hechscher- Ohlin model is the
model in which all factors of production move freely between sectors. In the model,
the reason country trade with each other is determined by the difference in ratio of
nation’s resources, and in technology, the combination of factors to produce a good.
H-O model explain that country will gain from trade if it exports product in whose
factor is well endowed and relatively intensive. Well endowed means that the relative
ratio of factor productions in country is higher than that of its partner countries while
relative intensive explicates that relative ratio of that factor to other factor used to
produce that good is higher than that of producing other good. H-O model provides
that international trade will impact on income distribution as a result of change
relative price of goods, and the relative factor price (Krugman, Obstfeld, & Melitz,
2012).
The theoretical models discussed above are based on the competitive market,
yet there are other types of markets such as monopoly, oligopoly, and monopolistic
competition. Krugman (1979) developed a trade-related theory based on economies of
scale in monopolistic competition market. The economies of scale is internal to firm,
that means firm get profit when increases its scale of output. Trading with other

countries leads to specialization, and created concentrated location of production. In
that case, countries can achieve cost advantage through mass- production which is not
based primarily on comparative advantage. The question put here is that how
countries decide which industries they will produce in large scale in order to compete
and export in world market. Krugman (1979) suggested that the market structure may
be a contributed factor. For instant, in oligopoly market, there are few suppliers for
whole market, and this is an appropriate condition for economies of scale. However,
the question arising is that how domestics firms can compete to foreign firms in world

Page | 7


market. Therefore, the reason that country trade is not only the differences in nation’s
resource and technology but also the economies of scale, and yet there is a
relationship between the Ricardian model, H-O model and economies of scale in
explaining the international trade pattern.
2.1.2 Free Trade Agreement
The standard theory of international trade is built on the assumption of trade
having no cost arise when deliver goods from exporting country to destination
country. Samuelson (1962), for instance, wrote a paper on analyzing the benefit from
free trade, and concluded that free trade potentially better for trading countries in
which winners can allocate benefit to the hurt; yet, the trade cost occur on
international trade and impact on the pattern of bilateral trade (Krugman, Obstfeld, &
Melitz, 2012). Anderson & van Wincoop (2004) analyzed the impact of trade cost on
trade flow between countries, and concluded that trade cost is greatly in trade between
countries. They are reasons for the existing of free trade agreement among trading
countries. The economic benefits of free trade are discussed in the international trade
theory, yet there is still political determinant of free trade between countries.
According to Krugman (1987), free trade agreement is a tool to prevent the trade war
between countries and interest group problem within country because free trade

provide certain rule and principle in trade agreed by both countries, and is difficult to
violate.

The issue posed is that how to approach a free trade in real world.

Krugman (1991) mentioned increasing trend of bilateral free trade agreement as
substitution for global free trade agreement. The paper indicates that free trade
agreement can benefit participants by reducing the consumer’s distortion, increase the
power in trade to other countries, and creating the efficient in production. Krugman
(1991) also mentioned the notion of ‘natural trading bloc’ which explains free trade
agreement as a natural process of countries located near together. Free trade
agreement also established between countries having similar economic characteristics
in order to increase its trading value (Baier & Bergstrand, 2004). World Trade
Organization defined free trade agreement is an accord between signed countries on
the significant reduction on trade tariff and other trade regulation.
Page | 8


2.1.3 The role of free trade agreement
2.1.3.1 Trade Creation and Trade Diversion Effect

Lipsey (1957) built a simple model with three countries A, B, and C. A is a
small country which produces wheat; B and C produce clothing. In an open economy,
A will import clothing from C because C’s relative price of clothing is lower than that
of B. Assume that A sets a tariff on clothing import from both B and C. If the tariff is
high enough, A will produce clothing herself. However, A is assumed to form a
custom union with B. As a result of those policies, price of clothing imported from B
without tariff is cheaper than C’s with tariff; and, A will import clothing from B. This
phenomenon is called trade diversion of A from C to B. Lipsey (1957) analyzed that a
customs union causing trade diversion by change the trading pattern in favor of

custom union suppliers. Lipsey (1960) mentioned that more likely the occurrence of
trade creation is the greater difference in the cost of production the same commodity
between two countries. Bhagwati (1971) clarified that a shift of importing goods from
lower cost supplier to higher cost one is called trade diversion. While Lipsey (1960)
advocate that trade-diverting customs union could not decrease the welfare of
importing countries because of the substitution in consumption, Bhagwati (1971)
pointed out substitution is not a sufficient conditions; yet, trade diversion effect can
reduce economic welfare under the assumption of fixed imported level. In details, the
change in import pattern due to formation of customs unions creates two opposite
effect on the welfare. The first one is the loss in term of trade. The second one is the
gain from consumption and production, which is the rise in trade flow among member
countries (Bhagwati, 1971)
After establishing a customs union, member countries can shift import from
higher-taxed suppliers to lower-no taxed suppliers; and domestic goods are substituted
by member’s low-cost goods. Johnson (1974) stated these above movements will
balance the increasing-welfare. Lipsey (1957) and Bhagwati (1971) defined the trade
diversion as the change in the locus of production from initial lower supplier to higher
member suppliers. Johnson (1974) argued that trade diversion can give a transfer from
locus of production which negatively impact on welfare, and create a substitution
effect which positively impact on welfare. Furthermore, Michaely (1976) asserted that
Page | 9


trade liberalization impacts on the trade pattern of signing countries in three ways.
First effect is the increase in new trade flow between bloc’s members. Secondly,
reducing trade barrier can divert the import from non-member suppliers to member
suppliers. Finally, the change in term of trade is an attributable to a rise in demand of
substitute commodity.
2.1.3.2 The intensive margin and extensive margin effect
In Krugman (1980), the pattern of trade is explained by two reasons. Firstly,

consumers in each country have different preferences on commodities with the low
substitution elasticity of foreign goods, so they demand the exporting goods. The
second reason is the economies of scale, so one country can produce goods in lower
cost than other countries in order to become least cost supplier in world trade. The
model of Krugman assumed that firms are the same, and can export to every country.
The only trading cost is transportation cost. As a result of those assumptions, the
impact of trade barrier on trade flow is explained by the change in level of exporting
of incumbent firms, which is called intensive margin. In details, if the trade barrier is
reduced, aggregate trade flow can increase because all firms raise their exporting
capacity of goods; so, every firm is equally impacted by trade policies changes. Melitz
(2003) stated that firms wanting to enter exporting market encounter not only the
variable costs such as transportation cost but also fixed cost which is independent to
the number of goods exported. As a result of that, Melitz (2003) built up a model to
find the static equilibrium of firms’ exporting in countries which expose to
international market, the paper concluded that trade liberalization effects on the trade
pattern of firm and reallocation resources within exporting industries through the
increase in number of importing countries, and the decrease in trade cost, both
variable cost and fixed cost. Firstly, the rise in trading countries is a condition to
increase in the productivity requirement for exporting firms. The least productive
firms can be forced to exit out of the market; the survival firms will increase its
volume of export to current and new destinations (Melitz, 2003). Secondly, the
decrease in trade cost can impact on trade pattern through the emergence of new
exporters. New exporter and current exporter capture the higher demand from export

Page | 10


market, and leads to an increase in volume of trade. In Ghironi and Melitz (2005), the
authors considered the trade cost as an exogenous factor of term of trade. The paper
concluded that the trade cost decreasing as a result of trade liberalization primarily

impacts on the change of trade volume through the extensive margin. The reason is
that countries will face a growth in number of firm exports, and decrease in
productivity breaking points.
Furthermore, Chaney (2008) supported the argument of Melitz (2003) by
imposing the intensive margins and extensive margins effect of trade cost on trade
flow. He pointed out that firm is different in productivity and suffers fixed cost in
trading besides variable cost. By adding new assumptions, Chaney (2008) stated that
change in trade flow due to trade barrier shock can be explained through two
mechanisms. The first mechanism is the intensive margin mentioned by Krugman
(1980). The second one is the extensive margin which is the change in the number of
exporter in different sectors. In general, trade barrier reduction such as free trade
agreement between two countries leads to the change in the number of good each firm
exports, and new exporters can enter the market. Crozet and Koenig (2010) confirmed
the theory in heterogeneous firms established in Chaney (2008). The paper
disintegrated effect of trade cost on trade flow in to intensive margin and extensive
margin as follow:

M  EXT  INT
j

j

(2.1)

j

Where M is the total trade cost elasticity of trading value
j

EXT is the trade cost elasticity of external margins

j

INT is the trade cost elasticity of internal margins
j

Crozet and Koenig (2010) concluded that the change in trade flow due to
fluctuation in trade cost varied across the industries. In details, the reducing in trade
barrier impacts on trade level greater in homogenous industries than more
heterogeneous industries. Extensive margin effect can dominate the intensive margin

Page | 11


effect in industry with high differentiated product. Moreover, Helpman, Melitz, and
Rubinstein (2008) stated that the profitability of firm export is higher if the importing
countries’ demand is higher, and lower trade costs. On the other hand, firm will not
export if the profit of firm is negative. That can explain the zero value in export
between two countries. He showed the primary bias in estimating impact of trade
barrier on trade flow is attributed to ignoring extensive margin effect. The reducing in
trade friction may not only lead to the expansion of trade between existent country
pairs, but also to create new trading partners.
From the above theoretical foundation, the impact of free trade agreement on
trade

flow

can

be

summarized

Page | 12

in

the

following

figur


Figure 2-1: The Relationship Between FTA and Trade Flow

+
FREE TRADE
AGREEMENT

TRADE BARRIERS

-

+

TRADE WITH
MEMBER TRADING
PARTNER

TRADE VOLUME OF

CURRENT
EXPORTING FIRMS

+
TOTAL
TRADE FLOW

+

NEW
EXPORTING FIRMS

-

+

-

-

TRADE WITH NONMEMBER TRADING
PARTNER

Source: Constructed by the Author

Page | 13


Figure 2-1 presents the theoretical foundation to answer research questions of
the thesis. The first one is that what the impact of free trade agreement on total trade

flow between signing countries is. Secondly, how that impact happens, or what
channels free trade agreement effect the change in trade flow.
FTA and Trade Barriers:
FTA by its objectives is to eliminate the tariffs, quota and other tariff
equivalent barriers among member states. The other purpose is to build an integrated
regional economy, so FTA will facilitate the reduction in other non-trade barrier such
as technical barrier, capital flow, cross border transit and human movement.
Trade Barriers and Trade with Member Countries and Non-member Countries:
According to trade creation and trade diversion theory, when the trade barriers
goes down, the trading cost also decreases both in variable cost and fixed cost. The
decrease in trade cost is a condition for two trading activities. The first one is the trade
diversion. Non-member countries suffer the full import tariff from a member country,
while other member countries are in favor of tax reduction. In details, the trading cost
of member partner is lower than that of non-member countries, so making member
trader is the least cost supplier. Therefore, there is the decrease in the trade flow
between member countries and non-member countries.
Trade with Member Countries and Exporting Firms:
The reducing trade cost between free trade countries will increase demand of
commodities export from member countries to other member countries. This leads to
the intensive margin and extensive margin effect in trade pattern. Firstly, the current
exporter will increase their volume of trade due to the increase in demand and lower
cost. The intensive margin effect will influence positively on trade flow between two
countries. However, this effect will be lower in more heterogeneous productivity
industries. Secondly, there will be new firms enter exporting market. Those new firms
will affect positively in the trade flow between countries. The more heterogeneous
productivity between firms the more positive effect of extensive margin on trade flow.

Page | 14

Exporting Firms and Total Trade Flow:
The intensive margin will impact positively on the level of goods existing firms
export, while the extensive margin will attract more firms come into the new
exporting market. Although the impacts of intensive and extensive margin are not as
the same level, in sector whose products substitution elasticity are high, extensive
margin impact will dominate intensive margin’s effect. However, both effects are to
increase the trade volume with the member countries.

Page | 15


2.2 EMPIRICAL LITERATURE
From the theoretical mechanism built in theoretical literature, this section aims
to provide empirical indicators used by previous researches. Aitken (1973) is the first
author which applied the theory of gravity model in order to estimate the effect of
European Economic Community (EEC) and European Free Trade Association
(EFTA) on the import and export of member countries. In his model, he proposed a
number of indicators for conceptions in the theoretical model. The dependent variable
was export value. The independent variables were Free trade agreement, Income of
exporting countries, Income of Importing countries, Distance between trading
partners. Those indicators have been used by latter researches on observing the impact
of FTA on trade flow of countries.
There are two main parts in this section. The first part is indicator for trade
flow, which is the dependent variable in this thesis. The second part is indicators for
independent variables which are Free Trade Agreement (FTA), exporter’s income,
importer’s income, and distance between trading countries.
2.2.1 Trade Flow
In empirical papers, there are three indicators for trade flow between countries.
Firstly, McCallum (1995) used the shipment value from one country to other country
as the representative for trade flow. In his paper, he analyzed the impact of trade

border effect in the trade between The United States of America (USA) and Canada.
The idea of research is that within-provincial trade of Canada is greater than the trade
with USA’s states, or foreign trade. By applying the gravity model as follow:

xij  a  byi  cy j  ddistij  eDUMMYij  uij
(2.2)
Where xij is the total export from region i to region j
Anderson and Wincoop (2003) examined the results obtaining from McCallum
(1995). The authors argued that the result is overestimated because it omitted trade
resistance variables which are called multilateral resistance terms (MRTs). Anderson
and Wincoop (2003) proposed a new estimated model as follow:
Page | 16


(2.3)
The model also used the export from countries i to j as the dependent variable
for trade flow. The differences with McCallum (1995) are the variables

, the

multilateral resistance terms. The meaning of MRTs is that it indicated the average
price of export country to the rest of the world’s price. The definitions of MRTs are as
follow:

(2.4)

(2.5)
Baier and Bergstrand (2007) conducted a research to estimate the impact of
FTA on trade flow. The paper used data of 96 countries around the world from 1960
to 2000. In his estimation model, the trade flow between i and j is the export value

from i to j divided by the GDP deflator to obtain real trade flow. Hur, Alba, and Park
(2010) applied the gravity model proposed by Baier and Bergstrand (2007) to find out
the hub-spoke effect of FTA using data on 96 countries around the world throughout
40 years from 1960. In the model, dependent variable is real exporting value, the
nominal exporting value divided by GDP deflator. However, Yang and MartinezZarzoso (2014) estimated whether ASEAN-China Free Trade agreement is trade
creation or trade diversion. In the paper, current export level in US dollar represent for
trade flow concept in international trade.
On the other hand, Jugurnath, Stewart, and Brook (2007) did not exploit the
export value as an indicator for trade flow. Instead of that, current import value is use
as a dependent variable in the model. The reason is that country often strictly reports
its import value for the tax purpose, so the import value is more correct than export
value.

Page | 17