View
3
Download
0
Category
Preview:
Citation preview
UNIVERSITÉ LIBRE DE BRUXELLES SOLVAY BRUSSELS SCHOOL OF ECONOMICS AND MANAGEMENT
ESSAYS ON THE EXPORT PERFORMACE OF VIETNAM
Vu Thi Hanh
Thèse présentée en vue de l’obtention du grade académique de
Docteur en Sciences Economiques et de Gestion,
sous la direction du Professeur Michele Cincera.
Academic year 2014-2015
ACKNOWLEDGEMENTS
This dissertation would not be possible without the guidance, comments and supports from promoters,
jury members of the defense, Vietnam Ministry of Education and Training (MOET), Vietnam
International Educational Development (VIED), my family, friends, colleagues and my employer, the
Foreign Trade University - Vietnam (FTU).
I would like to express my sincere thanks and my deep gratitude to Prof. Michele Cincera for all his
guidance during my research. His enthusiasm, encouragement, patience, sympathy, scholarly advice
and scientific approach have been extremely helpful. Without his help and counsel, the completion of
the dissertation would have been immeasurably more difficult. Indeed, whenever I encountered tricky
technical problems relating to the economic model, to the data treatment and analysis, he would always
raise questions which were in fact important clues enabling me to resolve them myself. I would like to
express my sincere thanks to my local promoter Prof. Tu Thuy Anh for her valuable comments as well
as her academic support along with Prof. Michele Cincera. Their dedication and support made a
significant contribution to the success of my research career.
My academic task can not be completed without the participation of jury members. I would take this
opportunity to extend my sincere appreciation to Prof. Bruno Van Pottelsberghe de la Potterie, Prof.
Laurent Gheeraert and Prof. Vu Bang Tam for their helpful comments, scientific advice, kindness and
time spent reading my dissertation in spite of their tough schedules.
I would like to express gratitude to my colleague working at the Foreign Trade University of Vietnam,
Doan Quang Hung for his contribution to my work in all discussions and interactions we had over the
years. I also want to thank my friend Vu Quoc Bao at the Department of Data Archives of the Vietnam
Customs Office for his support in data collection.
Furthermore, I am extremely grateful to the Ministry of Education and Training of Vietnam (MOET),
Vietnam International Educational Development (VIED) organization for their financial and
administrative support throughout my stay in Belgium. In particular, I wish to express my sincere
thanks to the ARES-CCD (Académie de Recherche et d'Enseignement supérieur - Commission de la
Coopération au Développement) for their generous support for the final stage of my research. Many
thanks to madam Do Thi Dieu Linh at VIED and madam Julie Sepulchre at the International Relations
Department of the Université Libre de Bruxelles for their essential administrative help.
My deep gratitude also goes to the Applied Economics Department (DULBEA) and the International
Centre for Innovation, Technology and Education Studies (iCite) of the Université Libre de Bruxelles
for providing me with the necessary facilities to work on my PhD dissertation. I also owe my deep
gratitude to iCite members who have patiently listened to my presentation, asked questions and made
comments which really helped me to improve my dissertation.
My special thanks to Jeffrey Silver for his help in revising the manuscript. I appreciate his kindness,
patience, enthusiasm as well as his time spent on helping me.
Finally, I would like to acknowledge with due gratitude, the support, and the shared love from my
family: my husband Dang Dinh Duc, my son Dang Duc Anh and my daughter Dang Dinh Hanh Kieu.
They are my endless enthusiasm, inspiration, support for which I try to work tirelessly. I am greatly
indebted to my parents who have taken care of me and encouraged me. I also remember with thanks
my parents-in-law to whom I owe my excellent husband for their support and inspiration during my
work on this PhD dissertation.
TABLE OF CONTENTS
CHAPTER 1. INTRODUCTION ............................................................................................................... 9 1.1. RELEVANCE OF THE TOPIC AND THE NECESSITY FOR SCIENTIFIC INVESTIGATION ......................................................... 10 1.2. MOTIVES FOR CHOOSING THE TOPIC ........................................................................................................................ 10 1.3. RESEARCH PROBLEMS AND QUESTIONS .................................................................................................................... 11 1.4. RESEARCH OBJECTIVES........................................................................................................................................... 12
1.4.1. General objectives of the research ................................................................................................................... 12 1.4.2. Specific objectives .......................................................................................................................................... 12
1.5. RESEARCH METHODOLOGY ..................................................................................................................................... 13 1.6. ORGANIZATION OF THE DISSERTATION ..................................................................................................................... 14
CHAPTER 2. CONCEPTUAL FRAMEWORK ON INTERNATIONAL TRADE ............................. 17 2.1. INTRODUCTION ...................................................................................................................................................... 18 2.2. THEORETICAL BACKGROUND .................................................................................................................................. 18
2.2.1. Classical trade theory ..................................................................................................................................... 18 2.2.1.1. Absolute advantage ................................................................................................................................................... 18 2.2.1.2. Comparative advantage .............................................................................................................................................. 19 2.2.1.3. Specialization of production and factor endowment ...................................................................................................... 21 2.2.1.4. International product life cycle theory .......................................................................................................................... 22
2.2.2. New trade theory ............................................................................................................................................ 24 2.2.2.1. Economies of scale and the pattern of trade .................................................................................................................. 24 2.2.2.2. National competitive advantage .................................................................................................................................. 25 2.2.2.3. Gravity model ........................................................................................................................................................... 26
2.3. EMPIRICAL FRAMEWORK ........................................................................................................................................ 27
CHAPTER 3. VIETNAMESE EXPORT STRUCTURE: PRODUCT DIVERSIFICATION
AND DESTINATION MARKET ................................................................................................................. 29 3.1. INTRODUCTION ...................................................................................................................................................... 30 3.2. TRADE LIBERALIZATION PROCESS AND ECONOMIC CONTEXT ..................................................................................... 30
3.2.1. Period 1986 - 1991 ......................................................................................................................................... 30 3.2.2. Period 1992 - 2003 ......................................................................................................................................... 32 3.2.3. Period 2003- 2012 .......................................................................................................................................... 35
3.3. EXPORT TRADE PATTERN ........................................................................................................................................ 37 3.3.1. Composition of export commodities ................................................................................................................. 37
3.3.1.1. General composition of commodities .......................................................................................................................... 37 3.3.1.2. Major export products, regions and economic groups ................................................................................................... 39
3.3.2. Export composition commodity by destination market ....................................................................................... 46 3.3.2.1. Sectoral dimension and country groups ....................................................................................................................... 46 3.3.2.2. Market and product concentration ............................................................................................................................... 50
CHAPTER 4. INTERNATIONAL EXPORT TRADE FLOWS OF VIETNAM: A GRAVITY
MODEL APPROACH ................................................................................................................................. 53 4.1. INTRODUCTION ...................................................................................................................................................... 54 4.2. THEORETICAL BACKGROUND ON THE GRAVITY MODEL ............................................................................................. 55 4.3. DATA AND METHODOLOGY ..................................................................................................................................... 62
4.3.1. Data description .......................................................................................................................................... 62 4.3.2. Model and methodology.................................................................................................................................. 63
4.4. EMPIRICAL RESULTS ............................................................................................................................................... 66 4.5. CONCLUSION ......................................................................................................................................................... 72
CHAPTER 5. THE ROLE OF FIRMS IN DETERMINING EXPORT TRADE PATTERN:
VIETNAM'S FOOTWEAR, RICE AND WOOD AND WOOD PRODUCTS SECTORS ........................ 74 5.1. INTRODUCTION ...................................................................................................................................................... 75 5.3. THE DYNAMICS OF FIRMS BY EXPORT CRITERIA ........................................................................................................ 80 5.4. LARGEST IMPORTING COUNTRIES BY EXPORT VALUE ................................................................................................ 82 5.5. METHODOLOGY AND DATA ..................................................................................................................................... 84
5.5.1. Gravity model and firms’ trade........................................................................................................................ 84 5.5.2. Data .............................................................................................................................................................. 86
5.7. EMPIRICAL RESULTS AND DISCUSSION ................................................................................................................. 90 5.7.1. Export value and the gravity model ............................................................................................................. 90
5.7.2. Export volume of firms and the gravity model ............................................................................................. 93 5.8. CONCLUSION ......................................................................................................................................................... 96
CHAPTER 6. DETERMINANTS OF FIRMS' EXPORTS: A CASE OF A DEVELOPING AND EMERGING
COUNTRY ................................................................................................................................................................... 97
6.1. INTRODUCTION ...................................................................................................................................................... 98 6.2. RELATED LITERATURE ............................................................................................................................................ 99 6.3. LOCATION OF VIETNAMESE FOOTWEAR, RICE AND WOOD AND WOOD PRODUCTS FIRMS ............................................ 103 6.4. METHODOLOGY AND DATA .................................................................................................................................. 106
6.4.1. Description of variables ................................................................................................................................ 106 6.4.2. Regression method ....................................................................................................................................... 108 6.4.3. Data ............................................................................................................................................................ 108
6.4.3.1. Data source and description ...................................................................................................................................... 108 6.4.3.2. Data treatment ......................................................................................................................................................... 110
6.5. EMPIRICAL RESULTS ............................................................................................................................................. 111 6.5.1. OLS regression results .................................................................................................................................. 111 6.5.2. Quantile regression results ............................................................................................................................ 113
6.5.2.1. Quantile regressions for footwear firms ..................................................................................................................... 113 6.5.2.2. Quantiles regression for rice firms ............................................................................................................................ 116 6.5.2.3. Quantiles regression for wood and wood products firms ............................................................................................. 118
6.6. CONCLUSION ....................................................................................................................................................... 120
CHAPTER 7. CONCLUSION .................................................................................................................... 121 7.1. MAJOR FINDINGS AND LESSONS RETRIEVED THROUGH THE EMPIRICAL ANALYSIS ............................................. 122 7.2. TO WHAT DEGREE DOES THE DISSERTATION ANSWER THE RESEARCH QUESTIONS AND WHAT ARE ITS LIMITATIONS? .... 124 7.3. IMPLICATIONS FOR VIETNAM'S EXPORT TRADE ....................................................................................................... 125 7.4. SOME PROPOSALS FOR FUTURE FURTHER RESEARCH ON EXPORT TRADE ................................................................... 127
REFERENCES ......................................................................................................................................... 129
APPENDICES .......................................................................................................................................... 136
ABBREVIATION
AFTA ASEAN Free Trade Area
APEC Asia-Pacific Economic Cooperation
ASEAN Association of Southeast Asian Nations
BTA Bilateral Trade Agreement
CARICOM Caribbean Community and Common Market
CEPII Centre d'Études Prospectives et d'Informations Internationales
CEPT Common Effective Preferential Tariff
CPI Consumer Purchasing Index
EU European Union
FDI Foreign Direct Investment
GDP Gross Domestic Product
GSO General Statistics Office
ILO
IPR
International Labor Organization
Intellectual Property Rights
KILM Key Indicators of the Labor Market
MFN
MRT
Most Favored Nation
Multilateral Resistance Term
OECD Organization for Economic Co-operation and Development
OPEC Organization of the Petroleum Exporting Countries
PPP Purchasing Power Parity
SITC Standard International Trade Classification
SME Small and Medium-sized Enterprises
SOE State-Owned Enterprises
TFP
TPP
Total Factor Productivity
Trans-Pacific Partnership
UNSD United Nations Statistics Division
VCO Vietnamese Customs Office
VCP
WDI
WITS
Vietnam Cummunist Party
World Development Indicators
World Integrated Trade Solution
WTO World Trade Organization
TABLES
Table 1. Illustration of the absolute advantage ..........................................................................................19
Table 2. Another illustration of the absolute advantage ............................................................................19
Table 3. Production cost ............................................................................................................................20
Table 4. Comparative advantage of country before trading ......................................................................20
Table 5. Comparative advantage of country after trading .........................................................................20
Table 6. Commodity composition of Vietnam's exports (1996-2000, millions USD and %) ......................39
Table 7. Major exports in each five year period (millions USD) ...............................................................41
Table 8. Major exports in period 2006-2012 (millions USD) ....................................................................42
Table 9. Gravity model for export flows of Vietnam ..................................................................................67
Table 10. Firm heterogeneity by sector (2006-2010) .................................................................................78
Table 11. Exporting firms'criteria across sectors ......................................................................................81
Table 12. Gravity factors and export margins (footwear firms) ................................................................87
Table 13. Gravity factors and export margins (rice firms) ........................................................................88
Table 14. Gravity factors and export margins (wood and wood products firms) ......................................89
Table 15. Export value of firms and gravity model ....................................................................................92
Table 16. Export volume of firms and gravity model .................................................................................95
Table 17. Determinants of export/revenue: OLS regression results ........................................................111
Table 18. Models of export intensity via OLS and quantile regression (footwear firms) ........................115
Table 19. Models of export intensity via OLS and quantile regression (rice firms) ................................117
Table 20. Models of export intensity via OLS and quantile regression (wood and wood products firms)
..................................................................................................................................................................119
FIGURES
Figure 1. Factor prices and good prices ............................................................................................ 22
Figure 2. International product life cycle ........................................................................................... 23
Figure 3. Porter's diamond ................................................................................................................ 25
Figure 4. Conceptual framework of the dissertation ........................................................................... 28
Figure 5. Evolution of Vietnamese trade and its country partners, 1986-2012 ..................................... 32
Figure 6. Historical trade foundation of Vietnam and its country group partners ........................ 33
Figure 7. Exports by country and country group (%) ...................................................................... 34
Figure 8. Vietnam's economic indicators ........................................................................................ 37
Figure 9. Share of export commodities, % of the total ................................................................... 37
Figure 10. Exports by geographical region .................................................................................... 44
Figure 11. Exports by country group .............................................................................................. 45
Figure 12. Export value of primary and manufactured products (millions USD) ................................. 47
Figure 13. Export value by country groups (millions USD) ................................................................. 47
Figure 14. Export value of primary products sector ....................................................................... 48
Figure 15. Export value of manufactured products sector ............................................................. 49
Figure 16. Vietnam's H-H market and product concentration index .................................................... 50
Figure 17. ASEAN countries' H-H market and product concentration index in 2012 .................... 51
Figure 18. 10 largest importing countries by export value (mil. USD) ................................................. 83
Figure 19. The correlation of firm's extensive and intensive margins ................................................... 85
Figure 20. Firm's factors in international trade .................................................................................. 99
Figure 21. Vietnam's number of firms across regions in 2008 ..................................................... 104
Figure 22. Vietnam's export value of firms across regions in 2008 ............................................. 105
9 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
CHAPTER 1. INTRODUCTION
SUMMARY
This chapter presents the context and scope of the dissertation as well as summing up major contents in
the following chapters: the theoretical background of trade and the origin of the two main branches of
international trade (chapter 2), a qualitative analysis helps to generalize the Vietnamese export structure
since the year of the country's political and economic reform (chapter 3), the export trade pattern of a
small and emerging country such as Vietnam replicating the gravity model (chapter 4), the role of
firms in determining the trade patterns corresponding to footwear, rice and wood and wood products
export sectors (chapter 5) as well as firms' internal factors affecting their export performance (chapter
6). This chapter also describes the general and specific objectives of the research, indicates the research
methodologies as well as the contribution of this research and concludes with the organization of the
dissertation.
10 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
1.1. Relevance of the topic and the necessity for scientific investigation
Trade liberalization and international economic integration are major and important issues especially to
developing countries including Vietnam. They have provided the country with many opportunities
such as foreign investment projects from developed countries, an increase in the State budget through
taxation on exports and imports, the higher level of employment, which have all contributed to
improving the standard living of the people. Since the country's reform (Doi Moi) in 1986, trade
liberalization has brought about an increase export value from 0.78 billion USD in 1986 to more than
72 billion USD in 2010 which accounted for an average of 70% increase to the GDP of Vietnam.
Moreover, Vietnamese firms are becoming increasingly aware of the importance of innovation and this
is becoming the main driver for Vietnam's deeper participation in regional and global economic
integration organization such as ASEAN1 and WTO
2.
Why does the dissertation focus on export trade? Firstly, exports reflect the country's role and position
in the global economy. The volume of exports shows the degree of a country' integration into the
global economy. Secondly, exports may contribute directly to the economic growth of a country and
export growth may lead to an increase in the demand for the country's products, thus helping to
increase the actual output. Thirdly, according to the new trade theory, increasing competition and
economies of scale lead to increased specialization in export products, which in turn helps to stimulate
the productivity level and workers' skills. As such, countries following the export-oriented strategy
may achieve sustainable development in a volatile global market.
1.2. Motives for choosing the topic
The contribution of Vietnam's economic reform in 1986 is considerable and paved the way for the
country to implement its trade liberalization policy and to establish trading relations with over 150
countries. However, existing studies on Vietnam's export trade only examine the country's trade with
certain countries inside or outside the Asian region. By investigating the Vietnamese trade pattern with
the world, we are able to track the precise trade trajectory of the country in several specific periods, which
also enables relevant trade policy recommendations to be formulated for policy makers. Moreover, there is
an absence of longitudinal macro and micro trade studies for an emerging developing country such as
Vietnam. In particular, in our model when analyzing international trade, it is important to take account
of sector and firm effects on exports.
Exports play an important role in Vietnamese trade, helping to increase the national revenue and to
improve the position of Vietnamese firms in the global value chain. Such an improvement of
Vietnam's position in the global value chain means being higher in the chain rather than carrying out
simple processing at the low-end. If the production factors which include labor, capital, and fixed assets
are analyzed in relation to firms' exports, plausible assessments of Vietnamese firms' contributions to the
global chain can be determined. It should be also noted that the country's trade policy reform has affected
1 Vietnam became a full member of ASEAN in 1995.
2 Vietnam became a full member of WTO in 2007.
11 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
the structure of Vietnam's exports during periods, has been subject to the effect of the volatile world's
trading market and the effect of an emerging giant exporter such as China.
To date, there have been some studies examining determinants of Vietnamese firms. However, the data
used in these studies came from the enterprise survey of GSO, which do not reflect the value of exports
of a firm in a specific year. Research questions addressed in these previous studies have identified
determinants of firms' decision to export. Data on total value of firms' exports have allowed the author
to measure the magnitude of change in the value of exports if a particular determinant changes. This
initiative aims to provide recommendation on export policy not only to the Vietnamese government
but also to firms when taking decision on exports.
1.3. Research problems and questions
In the age of globalization, interdependence among countries especially developed (North) and
developing countries (South) is increasing. While many large corporations in developed countries with
foreign investment projects are extending their market share and product life cycle, but on the contrary,
SME firms in developing countries are targeting major destination markets as a key strategy for export
promotion.
Vietnam's exports are facing some problems on international trade market for various reasons. Firstly,
the export structure continues to be mainly based on primary and labor-intensive products such as the
agriculture, forestry, fishery and footwear sectors. Secondly, low value added and labor intensive
products account for a large proportion of the total exports since almost all economic sectors of
Vietnam are involved in the assembly and simple processing stage of the global value chain. In
particular, the increasing share of the FDI sector in total exports may also influence sustainable
economic development of Vietnam because the advanced technology and management skill of these
firms cause fiercer competition for domestic firms. Furthermore, Vietnamese SOEs may be less
successful than non-state owned enterprises in export market.
We hope this dissertation can help strengthen the contribution of exports to the social and economic
development of Vietnam and constitute a reliable basis for formulating export trade policy. Firstly,
there is a need for greater efforts from Vietnamese government to implement social economic
structural reform with a special focus on SOEs, in order to strengthen investors' confidence. Secondly,
there is also a need to encourage the shift away from labor-intensive export products toward capital-
intensive products with special emphasis on innovation in particular through research and development
activities of firms.
At a macro level, the dissertation recommends that the Vietnamese government should support
exporting firms by actively seeking to sign trade agreements with strategic country partners such as
America, European and Asian countries as well as diversifying through the consumer markets of
African countries. The role of developing countries as potential importing partners can not be ignored.
For the agriculture sector, the Philippines is the largest market for Vietnam and African countries has
significantly increased their share of Vietnamese exports by volume.
12 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
For all the issues, the dissertation attempts to shed more light on modern trade theory by empirical
studies of Vietnamese export trade and addresses the following research questions:
- Since Vietnam implemented the economic reform in 1986, how has its export composition changed
at different periods?
- Although many studies show the role of GDP of importing countries in stimulating exports from
their country partners, it does not necessarily mean that high-income countries present an attractive
destination markets. The question of whether or not the wealthy importing countries encourage
Vietnamese exports in general and its firms’ exports in particular needs to be addressed.
- Does the economic integration of Vietnam into regional organization helps stimulate its exports?
- How does geographical distance (a proxy of transport cost) affect differently Vietnam’s exports and
its firms in the footwear, rice and wood and wood products sectors?
- Do these firms actually need hiring highly skilled workers and unrestricting fixed assets investment to
improve their export performance?
- How successful are state-owned firms in the international markets? Are they more effective exporters
than their non-state-owned counterparts?
- Is there any evidence to show that the older firms are more successful in export markets than the
young ones?
1.4. Research objectives
1.4.1. General objectives of the research
Generally, the research aims to analyze export activities of Vietnam through various periods since its
reform in 1986 and its economic integration into the regional and global economy. More importantly,
the dissertation also attempts to recommend specific export trade implications for policy makers as
well as providing guidance for exporting firms.
1.4.2. Specific objectives
Specifically, the dissertation aims to accomplish the following tasks:
- To carry out the qualitative analysis of Vietnam's export trade by destination market and principal
products.
- To find the association between Vietnam's export trade and its macro determinants where we are able
to identify the destination markets and the impact of the geographical characteristics on exports.
- To determine the relationship between firms' export trade value with firms' production factors such as
capital intensity, human capital intensity as well as firms' characteristics such as firm age, type of firm
ownership, firms' location, etc.
13 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
- To compare to what extent firms' factors affect differently their export performance depending on
some models. 1.5. Research methodology
The dissertation studies export trade of a single country namely Vietnam. It includes three empirical
study chapters investigating macro and micro determinants of Vietnamese exports which shed more
light on how theory of a country's factor endowment affect its export trade and on how gravity model
is used to explain trade pattern of a country. Vietnam, a small and emerging country in South East
Asia, has been implementing its economic reforms and trade liberalization for nearly 30 years.
Opening up its domestic market, it is an important tool for the country to benefit and to create trade
relationships with over 150 country partners both developed and developing. However, stimulus
factors for the country's trade may change for subjective and objective reasons. The effect of regional
and global economic integration on the export trade of country members may change from positive to
negative or vice versa particularly if a country is considering opening its market to a larger market
association.
The diversity of Vietnam's export products and the destination markets include almost all countries of
the world. We have used the synthesized database of the GSO, which demonstrates a structural change
of Vietnamese exports between 1986 and 2012. We applied a qualitative methodology which is helpful
in depicting a complete picture of Vietnam' exports. Moreover, for firms' exports, we also show
determinant factors in the table of descriptive statistics, which summarizes a sample of sub-populations
that the sample of three export sectors.
The qualitative method we applied in the dissertation was to carry out a literature review. In fact, our
contribution to the dissertation is to develop or continue a current branch of research. We show our
findings rather than replicating of previous authors' ideas. As such, this dissertation mentions the two
main strands of literature on the gravity model of export trade including theoretical and empirical
papers. This model criticizes and reviews the overall trade pattern of any country pair in so far as each
country differs from the other because of the effect of the two main factors the national income and the
transport cost proxy, which is the geographical distance.
Moreover, the gravity model is more detailed result at firm's data level so that the country's total
exports are broken down into firms' export value and volume which allow us to examine the role of
firms and sectors in determining the trade pattern. The dissertation also reviewed empirical papers,
which use the two main factors in our model the firm's performance and characteristics.
For quantitative methodology, our dissertation used export trade data at country and firm level to test
hypotheses for each determinant including firm and sector effects. Unlike the qualitative method, the
quantitative method allows us to draw conclusion about the whole population providing the sample of
sufficiently large. As stated above, we have extracted a dataset on Vietnam's export trade for the 1997-
2009 period. For firm's gravity model, we used export data of firms for the 2006-2010 period. We
analyzed determinants of firms' exports by combining firms' export data and the data of footwear, rice
and wood and wood products firms' characteristics for the year 2008. Since firms' export intensity
14 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
strictly lies in the unit interval, we transformed it into the type of logit then attempted with the OLS and
quantile regression methods.
1.6. Organization of the dissertation
The dissertation is organized as follows: Chapter 2 outlines the conceptual framework for international
trade in general and for export trade in particular. To be specific, we focus on both the theoretical and
empirical background. Firstly, for the theoretical background, the dissertation analyzes various trade
theory models in order to identify their strengths and weaknesses. Moreover, the development of
international trade theory is often linked to the practical changes in a country and to firms' trade.
Particular theory always needs to apply when making empirical study of trade. Moreover, the empirical
background examines various studies in which different indicators are used as dependent and
independent variables.
Chapter 3 gives an overview of Vietnamese exports for the 1986-2012 period. The chapter uses a pure
qualitative method to figure out the export structure of Vietnam for four sub-periods. The years 1986 to
1994 marked the period of "entrepreneurial policy-makers". In 1986 in particular, Vietnam vigorously
implemented its economic reform measures after 10 years of a centrally planned economy and the
country was faced with a severe economic crisis with 775% of hyperinflation in that year. Before 1986,
Vietnam mainly traded with socialist country partners such as the Soviet bloc, China and Cuba on the
basis of barter contracts. Exporting goods were mostly limited to raw materials and the payment was
based on the Ruble. During the period, important laws were passed such as in 1987 the new law on
Foreign Investment and the Land Law and in 1990, the Corporate Law and Private Enterprise Law.
This chapter also analyzes the pattern of Vietnam's exports for the second sub-period (1995-1999) of
regional economic integration and adoption of the market economy. Importantly, President Bill Clinton
lifted the U.S trade embargo against Vietnam in 1994 and one year later, the USA announced the
normalization of diplomatic relations between the two countries. During this period, Vietnam actively
participated in regional economic groups, becoming an official member of ASEAN in 1995 and of
APEC in 1998. Nonetheless, the country suffered from the turmoil of the Asian economic crisis in
1998. In the third period from 2000 to 2006, Vietnam continued to accelerate its economic integration
into the world economy by signing the BTA with the USA in 2001 and by joining the WTO in 2006.
In fact, that was also the year in which Vietnam' exports achieved their highest level since the
implementation of "Doi Moi". The creation of Vietnam's stock market toward the end of 2000 along
with a boom at the end of 2006 combined with the 150% increase in the VN-index's in particular,
proved a great challenge for the country. As the result of the world financial crisis in late 2008, the
effect of the contagion on the country was felt in 2009 with a sharp decrease in exports. The fourth
period (2006-2012) witnessed the collapse of many state-run conglomerates because of serious
problems ranging from poor efficiency and corruption to crony capitalism (Vuong, 2014). Notably,
many state run enterprises went bankrupt in this period, setting no little challenge for Vietnam's
national economy and its foreign trade.
15 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Chapter 4 describes relevant factors of the gravity model in order to explain Vietnam's export trade
pattern using data on Vietnam's exports with over 150 importing countries across five continents for
the 1997-2009 period. Although the traditional gravity model includes the two main factors, namely
the income of partner countries and the geographical distance, the dissertation develops additional
factors. Specifically, the chapter examined the effects of the importing countries' GDP per labor force,
relative factor endowment between country pairs and other proxy indicators for export transport costs.
As the chapter only dealt with export trade from Vietnam to its partner countries, its imports from its
partners in this case were not considered. Additional factors of gravity model were constructed based
on the available indicators. Firstly, we used the GDP per labor force as a proxy for the country's
productivity; secondly, the relative factor endowment is constructed by calculating the difference
between the GDP per capita of Vietnam and that of the importing country; thirdly, the exchange rate
between Vietnamese currency and an importing country’s currency can be a good proxy for the
multilateral resistance term. Moreover, the chapter examined the role of ASEAN in encouraging or
discouraging exports. In this chapter, the fixed effect model is consequently preferable. However, the
fixed effect regression model does not report the time-invariant variables, therefore the random-effect
model can be used as an appropriate reference. In general, the regression results from the coefficients
were as expected except where the geographical distance was insignificantly negative which suggests
this trade cost should be seen in the context of Vietnam's simultaneous trade with the rest of the world.
Chapter 5 examines the effects of gravity factors on firms' exports in footwear, rice and wood and
wood products sectors for the 2006-2010 period. We have treated firms who export to different partner
countries as a group then attempted to examine how each firm-country group performs differently in
continuous years of the period. Surprisingly, we found that gravity factors such as the GDP of the
importing countries and geographical distance between Vietnam and importing countries do not give
the same effects compared with those from gravity model using aggregated country level data. More
importantly, our findings in this chapter confirm the role of individual firms in determining their export
trade pattern.
Chapter 6 contains the findings of the analysis obtained from the quantitative study of three export
sectors such as wood and wood products, footwear and rice, based on various econometric models. We
identified the relationship between those determinants and a firm's export intensity. We are especially
interested in identifying the effect of the factors namely human capital intensity, firm’s size, capital
intensity and type of firm ownership (state-owned firm may be a supportive factor of exports).
Although highly qualified workers have been considered an advantage for Vietnam in attracting FDI
capital (Nguyen and Nguyen, 2007), they are becoming a disadvantage for the country's exports
because the export sectors require mostly low-cost labor in order to increase export volume. Arising
proportion of firm's exports to total sales is a good indicator that decision has been taken to increase
exports. The chapter also analyzed the role of firm location in improving exports as well as firm
distribution across the country since the geographical location, to some extent influences the exporting
activity of firms particularly those producing agricultural products. The chapter also compares the
export value of the three sectors and examines the dynamics of firms through their annual number of
contracts and their export value.
16 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Chapter 7 concludes the dissertation with some concluding remarks and makes important suggestions
on important points of Vietnam's export trade policy. Since the determinants have different effects on
Vietnamese exports depending on the sector, therefore our policy recommendations focus on a single
export sector. We study macro and micro level export determinants so as to provide guidance for both
policy makers and firms seeking a method to improve their export performance. Solving the export
problem is a major concern of firms seeking to strengthen their international market position.
17 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
CHAPTER 2. CONCEPTUAL FRAMEWORK ON INTERNATIONAL TRADE
SUMMARY
This chapter analyzes the evolution of international trade theory beginning with a logical explanation
of the methods countries used to exchange goods. Attitudes to the reasons for international goods
exchange depend on a country's production innovation. However, classical trade models such as
absolute and comparative advantage trade theories explain trade among countries using a simple
example of any two countries. Modern trade theory focuses on trade leverage rather than the available
advantages of a country such as natural resources or low cost of labor. We also mentioned and
criticized the role of firms in international trade as a core factor in determining a country' trade pattern.
Determinant factors affecting trade performance include both internal and external factors of the
country and the firm.
18 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
2.1. Introduction
Theoretical and empirical background section provides a critical map for the dissertation on Vietnam
trade pattern with its partner countries. In fact, the term export trade applies not only to a country pair
but also to a country with several economic blocks. Moreover, smaller entities such as firms play an
important role in influencing trade pattern since they are dynamic and different firms have different
potential and competitive strategies.
The first section of this chapter discusses various theoretical trade theorems. Firstly, classical trade
theory describes a country's trade using pure models in which two countries enter the trade model.
Secondly, new trade theory discusses production factors of a country using a data set of many countries
to prove the validity of theoretical models.
The second section of this chapter considers a wide range of empirical papers served as conceptual
framework for country-to-country trade and firm-to-firm trade. More importantly, this section builds
on the theoretical framework discussed in the first section and describes a conceptual framework
employed in the dissertation for the empirical analysis of the research question under investigation. The
conceptual framework covers three main studies including both quantitative and qualitative methods.
The qualitative study describes the general picture of Vietnamese export trade for a long period of
1986-2012. The qualitative study focuses on answering typical research questions including what to
export, when to export, why to export and for whom to export. The first quantitative study investigates
the export trade pattern of Vietnam, the second examines the role of individual firms in determining
their export trade pattern and the third one attempted to find the export determinants.
This chapter aims to show that the dissertation is a further contribution to the series of research papers
on international trade and is only one to analyze the example of Vietnam, an emerging and developing
country. The theoretical perspective itself gives context for our three empirical studies of Vietnam's
export trade. In turn, the empirical studies confirm and shed more light on the theory.
2.2. Theoretical background
2.2.1. Classical trade theory
2.2.1.1. Absolute advantage
International trade theory has long been proposed as a mean to help develop international market for
countries seeking more consumer markets for the goods and services produced within their own
borders. The principle of the absolute advantage theory was first introduced by Smith (1776) to
explain that any individual or country that can not produce one good more efficiently than others will
tend to purchase that good from one of them. Where that efficient producer and consumer do not live
in the same country, the principle is applied if no monopoly is imposed on the home country. However,
he pointed out that where a domestic product can be purchased as cheaply as a foreign equivalent product,
such regulation would clearly be pointless even though generally undesirable. The productivity of labor in
the tables below illustrates the absolute advantage.
19 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Table 1. Illustration of the absolute advantage
Produce Country A Country B
Commodity 1 (kg/capita/hour) 1 3
Commodity 2 (meter/capita/hour) 4 2
In another case, since the tradable price of one good is the total costs, the absolute advantage can be
otherwise expressed as
Table 2. Another illustration of the absolute advantage
Produce Country A Country B
Commodity 1 (USD/kg) 1 3
Commodity 2 (USD/meter) 4 2
The absolute advantage theory inspired trade liberalization and was critical of the disadvantages of the
mercantile system, which had been earlier adopted by European governments since that system
allowed governments to increase a nation's wealth by regulating all the nation's commercial interests.
The mercantile system in some cases encouraged a nation to export rather than to import certain
commodities. Exports of certain manufacturing input materials were discouraged to enable finished
products to undercut products of other nation. It also encourages the import of manufacturing input
materials so that the workers produce finished goods more cheaply than other nations.
Unlike the mercantile system, which describes the gain in wealth of one nation as the loss of another,
the absolute advantage theory underestimates the role of government in intervening a country's trade.
However, the theory does propose that one country should specialize in producing a good that the other
country can not. It also emphasizes the role of individual firms in a free trade system. However, this
theory does have some limitations. Firstly, it does not include the case where a country has not
international trade advantages while others may. Such cases have attracted more attention in recent
years in context of North-South trade. Secondly, the theory also considers labor as a mere input
production factor in the international distribution of labor and does not take account of the proportion
of labor in different goods from different countries.
2.2.1.2. Comparative advantage
The following Ricardian model (Ricardo, 1817) developed the comparative advantage theory, where
opportunity cost is described as a reason for one country to specialize in what it can do best. This
principle partly explains one limitation of the absolute advantage theory whereby any country can act
in its own interest when engaging with the international market. A country can benefit through
specialization to produce one or more goods at a comparatively lower price than a rival country. In
turn, a country producing the good less efficiently (at a higher price) than the other can import that
good. It is important to see an individual country's gain in the context of the total sum of all goods
traded. If in fact, a country is unable to produce a particular good more efficiently than a rival country,
it can still outperform its rival with the production of other good.
20 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Table 3. Production cost
Product Country A (wage hour) Country B (wage hour)
One unit of wheat 15 10
One unit of wine 30 15
In this example, country B has an absolute advantage over country A in producing both wheat and
wine. According to the explanation of Smith (1776), country B may not find it advantageous to import
either good from country A. However, under comparative advantage theory, country B could benefit in
another way. Since one unit of wine produced in country A is equivalent to the costs of two units of
wheat in country B, while the cost of one unit of wine produced in country B is equivalent to
production cost of 1.5 unit cost of wheat in country A, therefore, the production cost of wine in country
B is comparatively lower than in country A. Likewise, in country A, the production cost of wheat is
comparatively lower than in country B, which means that country A has a comparative advantage for
wheat and country B has a comparative advantage for wine. In the absence of international trade, the
two countries would themselves produce both goods.
Table 4. Comparative advantage of country before trading
Country Wheat unit quantity Wine unit quantity
A 8 5
B 9 6
Sum 17 11
Assuming that country A and country B merely specialize in producing wine and wheat and exchange
these products then following quantities would be achieved:
Table 5. Comparative advantage of country after trading
Country Wheat unit quantity Wine unit quantity
A 18 0
B 0 12
Sum 18 12
It is therefore clear that if each country focuses on producing a good in which it has a comparative
advantage, the total quantity of wheat and wine would increase after trading between the two countries.
However, as with the absolute advantage theory, there are certain limitations in the comparative
advantage theory in some circumstances:
- Assuming identical production cost of goods in two countries A and B and high transport cost
between them compared with a country C which shares a common border with an export country A,
the exporting country A may decide not to trade with the distant country B.
- In some cases, a country is able to benefit from economies of scale so that the production cost of a
good accordingly falls because the fixed cost is spread over more units of output. In addition to the
labor cost, other cost advantages are considered which also therefore affect the choice of the partner
country.
21 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
- Since different countries do not necessarily produce the identical quality of a nominally similar
product, this may affect the choice of partner country.
- Assuming there are neither export and import taxes nor trade barriers between the two countries.
- It is also assumed that perfect information on a good is easily accessible to both the seller and the
buyer in the international market.
2.2.1.3. Specialization of production and factor endowment
Factor endowment in production originally developed by Heckscher (1919) and Ohlin (1933)
describes a given quantity of factors of production within a country rather than the dynamic changes in
those factors. While comparative theory mentioned production factors including labor, land and capital
as essential resources of a country's advantage, the analysis of resources distribution is based on
assumption. In actual fact, one country can be intrinsically different from another in its production
endowments. For example, Vietnam tends to export more agriculture and seafood products because of
the favorable factor endowments climate and geography (the two large plains Red River Delta and
Mekong River Delta). A long coastline is also an advantage for the country in fishery product sector.
Britain in contrast has the advantage of several key sectors such as aerospace, pharmaceuticals,
automotive sector thanks to its skilled labors.
According to the Heckscher-Ohlin theorem, a country will tend to export goods that use its abundant
factors intensively and to import goods that use its scarce factors intensively. The theory proposes the
important assumption that the marginal rate of substitution is independent of the scale of consumption.
Specifically, the consumption reduction for one good is the additional consumption amount for another
good if the level of utility of a person remains unchanged. Since resources are not unlimited and the
manufacturer has to determine the proportion of labor and capital allocated among goods, resources
available for distribution are therefore inevitably restricted. A manufacturer has to use his own capital
to pay for labors and fixed costs and the final cost of a product needs to cover these basic costs and
ultimately determines the price of the goods. Leaving aside other relevant price factors, this will give
the normal price conventionally called the unit price.
Figure 1 shows the relationship between the relative price of cloth together with the price of food and
the wage rent ratio. As labor becomes more expensive relative to capital, cloth, which is labor-intensive
in production, finds itself at an advantage and becomes relatively expensive compared to food.
And both home and foreign country use the same technologies, the same FPGP curve is applicable to
both countries and under free trade, the relative price of cloth will be the same in both countries.
Therefore, the wage-rent ratio will also be the same in both countries.
In an autarky economy, according to the Heckscher and Ohlin theorem, the labor-intensive good would
be relatively cheaper in the labor-abundant country therefore under free trade, the labor-intensive good
would be exported by the labor-abundant country and the capital-intensive good would be exported by
the capital-abundant country.
22 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Figure 1. Factor prices and good prices
Relative price FPGP
of cloth, Pc/Pf
17
5 Wage rent ratio, w/r
Source: Krugman, Obstfeld and Melitz (2012)
Nevertheless, the Heckscher-Ohlin theory reveals some limitations arising from its fundamental
assumptions. Firstly, the restrictive model examines the case of only two countries engaging in trade
with two goods and two production factors. Secondly, the model is static since does not cover the case
of two countries trading over a period of time therefore, the effect of any changes in production factors
such as capital and labor on country trade are not considered. Thirdly, consumer's demand, which also
influences the factor price, is ignored in the model. Fourthly, Leontief (1953) has countered the H-O
theory by presenting the case of American trade where he found that America exports labor-intensive
goods and imports capital intensive goods whereas America is in fact a capital abundant country.
Fifthly, just as the absolute and comparative advantage theory, the H-O theorem neglects other
important factors influencing commodity prices and international trade such as technology, natural
factors and differences in quality of labor.
2.2.1.4. International product life cycle theory
Vernon (1966) developed the theory of the international product life cycle linking international trade
with investment based on a product life cycle by using a database of a capital-intensive industry in the
United States.
Using the pure qualitative research method, he proposed three-main sub-stages of a product life cycle,
which influence differently the decision on production costs of a firm. At the first stage when a new
product is introduced, producer is more likely to be aware of the importance of more effective
communication with the potential market than of product innovation. Therefore, a producer is more
likely to choose his local country as production location.
At the introduction stage, a prospective producer tends to make the location choice based on the initial
capital requirement since new producer is very much aware of a potential risk arising from the fact that
the product is not yet well known and competitors have a far stronger reputation and far larger market
share. For those reasons, prospective producers are more likely to choose the local country to locate the
23 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
firm whereas existing producers are more easily able to expand their production overseas and benefit
from economies of scale.
Maturity stage comes as demand particularly from foreign countries for a product expand which leads
to existing producers starting to export. Depending on the distance between exporting and importing
country, export and import tax on products, producers gradually come to realize that such costs do
matter especially cost of labor and raw material, which improve their competitiveness and thereby
make it easier for them to serve foreign markets. On the other hand, other domestic producers of
similar product may become more competitive where local demand may seem to contract.
Figure 2. International product life cycle
Stage 1 Stage 2 Stage 3
Introduction Maturity Standardization
Highly skilled Largely unskilled production workers
0 10 20 30
Time (years)
Source: Vernon (1966)
At this stage, firms may therefore find it necessary to seek new markets. Vernon (1966) was critical of
the fact that as long as the marginal production cost plus the transport cost of the goods exported from
the United States was lower than the average cost of prospective production in the import market, a
producer in the United States would prefer not to make investment on that market.
Product standardization is the advanced stage of a product life cycle. Local consumers in developed
countries are very familiar with the product whereas developing countries have high demand for the
products and offer attractive labor costs and may offer a low-cost captive source of supply of raw
material. In this regard, less-developed countries may supply cheap labor cost and perspective source
of material as well. At this stage, producers also need to consider creating trademark. Certain suppliers
of food and drink services in particular, may start thinking about the need to grant franchises, which
makes market communication of vital importance. There are various types of investment at certain
sub-stages of product development depending on the domestic and export market status. The
investment decisions of local producers are closely linked to the trade policy decided by the
government of developing countries for example import substitution or export oriented. If a developing
country applies import substitution policy, investors may face restrictions including import duties,
import licenses or currency transactions. In contrast, developing country following an export-oriented
policy to accelerate their industrialization may offer favorable terms such as reduced tariff barriers for
Labor
24 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
imported raw materials or floating exchange rate to potential foreign investors seeking to relocate their
manufacturing facilities. Nevertheless, investment stimulation is not always suitable for all developing
countries because investment by a highly competitive longer established foreign producer may damage
a developing country' nascent industries. Moreover, a floating exchange rate regime is an inherent risk
for emerging countries because if a local currency is anchored to the USD, the wave of high liability
dollarization generated may result in financial fragility.
The international product life cycle like many other theories has its own limitations in that all the sub-
stages of one cycle are purely hypothetical depending on observing one specific economy of the US.
Investment decisions rely on the investment demand of the US producer and the changing consumer
markets. Furthermore, the model does not support the use of historical data for forecasting. The life
cycle of a product depends on consumer sale. Not all consumers buy in the introductory stage. Some
people buy early, others buy some time later than their friends. In particular, the product life cycle may
not synchronize with that of a similar product in a particular developing country or conversely, other
developed country competitors may already have started to invest in the same country, which may
constitute an obstacle for the first producer in deciding whether to invest.
2.2.2. New trade theory
2.2.2.1. Economies of scale and the pattern of trade
Krugman (1980) was the first to recognize that economies of scale is a prime cause of international
trade and asserted that countries with the relatively large domestic market are more likely to export
such goods. Although his theory also acknowledges labor as an important factor in the production to
explain why a country exports, its approach is very different from comparative cost theory. Whereas in
the comparative advantage model, labor is one factor in determining the product value, in the Krugman
theory, he assumes it is a function of a country's output of one good. Importantly, under his theory,
output of one good is equal to the sum of individual consumptions. Krugman has also assumed that
there is no free entry and exit of a firm otherwise, the profit margin would always be zero.
Krugman examined equilibrium between firm's output and wages in the context of both a closed and
open economy. By combining the utility function with the labor factor in a cost function, he concluded
that firms costlessly differentiate their products because two firms never want to produce the identical
product. If there are a large number of different goods being produced, the pricing decision of any one
firm will have negligible effect on its marginal utility of income. Especially where profits are positive,
new firms will enter the market, causing the marginal utility of income to rise and profits to fall until
profits are driven down to zero. Assuming zero transport cost, trade would not affect the scale
production and the gains from trade would come solely through increased product diversity.
In the open trade system, assuming zero transport cost, trade would occur in the presence of increasing
return that is gains from trade would occur if the world economy were to produce a greater diversity of
goods offering each individual a wider choice. Even if we include transport cost in this equilibrium, the
elasticity of export demand would be the same as that of domestic demand indicating that transport
25 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
costs would have no effect on the number of firms or output per firm. However, transport costs would
matter if wages were assumed not to be equal in the two countries. Firms would be more likely to
concentrate production near their largest markets so that they could minimize their transport costs.
Each country would be a net exporter to its largest market. However, wages would not normally be
equal particularly in smaller economies. Whichever country has a larger market for a product would be
a net exporter of that product and a net importer of another product.
2.2.2.2. National competitive advantage
Porter (1990) asserted that differences in national values, culture, economic structure, institutions and
histories contribute to competitive success. Since nations can not be competitive in all fields and
industries, they need to develop their own appropriate strategy so as to become more competitive in the
international market. As such, he proposed a diamond model, which highlights the importance firms
and other factors in promoting country's competitive advantages. Using the case study method, Porter
conducted a four-year study of 10 major trading countries and postulated interesting findings on how a
country can improve its competitiveness. According to this model, four main attributes create
determinants of a national competitive advantage.
- Factor endowments: Porter criticized standard economic theories especially those of Adam Smith and
David Ricardo in their economic models who postulated certain key production factors such as land,
labor, capital and natural resources.
As analyzed in the previous sections on the classical trade theory, any country endowed with one or
more factors can gain an advantage on the international market and a labor-intensive country tends to
import capital-intensive products whereas capital-intensive country tends to import labor-intensive
products.
Figure 3. Porter's diamond
Source: Porter (1990)
Firm strategy,
Structure, and Rivalry
Demand conditions Factor endowments
Related and Supporting
Industries
26 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
This leaves open the question of how a country without low cost labor or natural resources can raise its
competitiveness. Does a country faced with tough competition from other nations learn to respond to
challenges? Porter indicates that firms manage to find opportunity even in the absence of natural
endowments. Such firms need to upgrade their labor productivity and to innovate.
- Demand conditions: Home-demand to some extent reflects the needs of emerging buyer and
encourages domestic firms to innovate and improve standard quality of their products. In some cases,
sufficiently large domestic demand can lead to increased demand from other nations. In other words,
local demands may significantly affect global consumption.
- Related and supporting industries: Manufacturing a finished product involves different stages in the
production chain, which need to be linked. Within a single country, a firm may not be able to cover all
the production stages and hence may need to cooperate with another firm, which can provide it with
support in the form of input materials or design work. For example, a shoe manufacturer needs to
cooperate with a leather producer as a supplier of an essential input material. In the context of the
global value chain, firms from one country can cover one important link in the chain and thus a
finished product is international although domestic supporting industry where available can help a
manufacturer to reduce costs and save time since they are easier to approach and better known to him
than in third country's supporting industries. Where a firm can not cover all stages of production then it
is very important to have a strong supporting industry. Furthermore, supporting industries can provide
innovation and upgrading for a manufacturer and therefore improve technical quality.
- Firm strategy, structure and rivalry: Firms' strategy and structure often differ from one country to
another. These differences depend on the national context as well as on the types of sectors. In some
countries, small and medium firms are more productive than larger ones while in some other countries,
firms are strictly hierarchical in organization and management practice. Even in any sector of a
country, there may be both domestic and foreign competitor supplying similar product. Globally, the
competition is fierce especially for firms entering the market. However, competition also motivates
firms to innovate in various ways such as improving product quality, introducing new products, which
would drive less innovative firms to exit the market. Competition in some respect is to be considered as
an advantage since it encourages firm to improve their performance.
2.2.2.3. Gravity model
The gravity model first proposed by Tinbergen (1962) to explain the pattern of trade by one country is
considered to be analogous to the Newton’s gravitational law. According to the basic model, the size of
the economy of exporting and importing country is the attractive factor in determining a country's trade
pattern whereas the geographical distance between a country pair is the unattractive factor. In many
studies on trade prior to ours using the gravity model, the attractive factor is not limited to the income
of country for which GDP is a good proxy but also includes other indicators such as GDP per labor
force, population, the relative factor endowment. Furthermore, variable "cost" is not limited only to the
geographical distance but includes other factors such as social distance, whether the country is
27 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
landlocked, whether it is a member of a common economic group, whether it shares a common
language with its partners or whether one country is colony of the other.
Theoretical foundations of the gravity model mostly regards trade volume between a country pair as a
function of supply and demand factor in the supply and demand equation. Although the gravity model
explains the trade pattern between a country pair and most literature uses data on trade at country level,
the role of the firms in determining trade is also of crucial importance since firms are dynamic and
firms active in the international market are more competitive than firms only active in the local
markets.
2.3. Empirical framework
The contribution of exports to the country's economy has been analyzed in the above sections.
However, the product structure of Vietnam's exports still needs to be further addressed. The description
of a country's export product structure also indicates one country's competitive capacity in a particular
sector. The fact that developed and developing countries have different product ranges indicates they
have different resources.
Since Vietnam's economic reform almost three decades ago when it opened its market to foreign
investors, its export product structure and trade partners have fundamentally changed. Vietnam's
regional and global economic integration is which enables local firms to penetrate very distant markets
including the highly dynamic market such as the United States of America, etc. With over 150 export
partners, the export volume has continuously increased. Our dissertation gives a general description of
Vietnam's economic integration through exports for the 1986-2012 period. Since Doi Moi in 1986,
Vietnam has become not only an official member of the regional integration group ASEAN but also of
the WTO, as well as concluding many bilateral trade agreements with countries all over the world.
Specifically, our dissertation attempts to analyze Vietnam's export trade pattern at the macro and micro
level by applying the gravity model to exports data at country and firm levels. Of course, a country's
exports are the aggregate value of the individual firms' exports in a given period. However, export
firms' dynamics vary which directly affect their comparative export performance. Our dissertation also
attempts to analyze the role of the firm's extensive and intensive margins in determining Vietnam's
trade pattern in three export sectors. Furthermore, firms active in different export sectors have different
characteristics and therefore require appropriate distribution of production resources.
Internal factors of firm in a particular industry are determinants, which may encourage or discourage
firm's export performance. Figure 4 below shows the framework of the dissertation and presents
Vietnam' exports performance since the year of country's economic reform 1986. Four important
chapters of the dissertation show the evolution of Vietnam's exports as well as the relevant macro and
micro determinants. One chapter uses a purely qualitative method while the other three chapters use
quantitative methods based on the three main data sources including the UN Comtrade, the GSO and
the VCO of Vietnam.
28 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Figure 4. Conceptual framework of the dissertation
Source: Author's research
Chapter 5
Firms to country
exports
(Quantitative)
Chapter 3
Export structure: Product diversification and destination market
(Qualitative)
Chapter 4
Country to
country exports
(Quantitative)
Export
entity
* GDP
* GDP per labor
* Difference in GDP per capita
* Distance
* Trade block: ASEAN
* Landlocked country
*Common border
*Colony
* Economies of
scale
* Types of
ownership
* Experience
* Location
* Capital intensity
Chapter 6
Exporting firms and
their internal
determinants
(Quantitative)
29 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
CHAPTER 3. VIETNAMESE EXPORT STRUCTURE: PRODUCT DIVERSIFICATION
AND DESTINATION MARKET
SUMMARY
This chapter gives an overview of Vietnam's export trade since its economic reforms in 1986 using
only the qualitative method. Vietnam's export pattern is investigated against the background of
Vietnam's economic reforms in three main periods. The 1992-2003 period in particular represents a
major landmark in Vietnam's economic and trade integration into many groups such as ASEAN and
APEC. When Vietnam became a full member of the WTO in 2006 that was a signal that the country
had completely integrated into the global trading structure. This chapter also analyzes Vietnam's export
structure, which had been changing dramatically in the course of the three periods. Indeed, export
revenue from manufactured goods has gradually exceeded that of primary goods showing that the
country has been increasingly concentrating on developing products requiring skilled labor. The
chapter also analyzes Vietnam's export markets as well as the contribution of the most dynamic
importing countries and country groups including the USA, the EU, Japan and China.
30 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
3.1. Introduction
The break-up of the Soviet Union and the fall of communism in the Central and Eastern Europe bloc in
the 1980s had negative consequences for other communist countries in South Eastern Asia including
Vietnam. The country was faced with the dilemma when the western countries did not open up their
economic relationship with Vietnam but followed the embargo policy of the United States of America.
The shortcomings of the centrally planned and closed economy pushed Vietnam to the brink of
bankruptcy. The country was also facing a major economic crisis in 1986 when the highest rate of
hyperinflation reached 775% and there was still three-digit inflation for a further two years.
For ten years following the country's independence in 1975 until the economic reforms in 1986,
Vietnam's international trade relations were mainly confined to the countries of Socialist Republic bloc
and covered some capitalist countries including Japan, France, Sweden and India. Meanwhile, export
products were mainly manufactured from various natural resources such as coal, chromium, cement
and some agricultural products. Foreign trade payment was made in Ruble, which was then converted
into US dollar. In 1985, the country's export trade with the Socialist Republic countries bloc earned
425.8 million Rubles and 272.7 million Rubles were earned through trade with some capitalist
countries.
The year 1986 when the leaders of Vietnam's VCP launched a series of economic renovations
including the organizational and personnel structure, administrative system, economic and political
system is considered as a turning point of the Vietnamese economy. Specifically, these fundamental
changes applied to the country's economy constitute a transition from a centrally planned economy to a
socialist-oriented market economy. The opened position of the VCP on the country's economy is
shown by the Government's acceptance of a multi-sector economy including the state-owned sector,
collective sector, domestic private-capitalist sector, foreign-owned sector. Following the VCP
guideline of 1986 whereby Vietnam was to aim to be on friendly term with the countries of the world,
since then it has started to integrate into the world economy.
Some economic indicators such as GDP, trade revenue, the FDI attraction index confirm how much
Vietnam's economy has been transformed since 1986. From 1986 to 2012, the contribution of exports
in GDP rose dramatically peaking at 73.5% in 2012. In particular, export revenue in 2012 reached
114.5 billion USD, which corresponds to 145 times the country's export revenue in 1986. Moreover,
the number of Vietnam's importing partners also rose annually from 33 in 1986 to a peak of 220 in
2002.
3.2. Trade liberalization process and economic context
3.2.1. Period 1986 - 1991
Vietnam's economic crisis was mainly caused by a misguided economic policy which forced its leaders
to launch Doi Moi, changing their political approach and taking new economic and trade initiatives. A
lengthy virtually closed economy had many consequences for the economic approach and its development
strategy of Vietnam. After Doi Moi, the economy went through a long recession before recovering from
31 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
the introduction of the market-oriented economy. Negative economic indexes pushed back the national
economy leaving it far behind the world economy.
The economic reforms in 1986 were a turning point showing that Vietnam was ready to open up to the
countries of the world. The first example of the Government's new direction was the 1987 first law on
Foreign Investment in which the Socialist Republic of Vietnam welcomed and encouraged foreign
organizations and individuals to bring capital and technology to Vietnam on the basis of mutual
benefits and nondiscrimination between domestic and foreign investors. Another example was the
1990 Cooperate and Private Enterprise law whereby all Vietnamese entitled citizens were allowed to
establish private enterprises. Although private economic entities had been legally recognized in 1972,
many of those were subsequently nationalized and converted into SOEs. This economic decision badly
affected the growth of the economy and was one of the main causes of the 1986 national crisis.
Furthermore, the 1990 major legal milestone was an importance factor in encouraging both the private
and state-owned enterprises to boost their contribution to the growth of the economy. However, the
1990 law was also seen as an important step forward compared with the 1972 Commercial Code,
which was restricted to only five forms of enterprises: partnerships, simple share capital companies,
joint capital companies, limited liability companies and share holding companies.
As early as three years after Doi Moi, Vietnam has made remarkable progress including a reduction in
the rate of inflation to 67.4% and whereas in 1988, the country had to import nearly half a million ton
of rice, a mere two years later, it was able to export over 1 million ton. The Foreign Direct Investment
law of 1987 not only encouraged an increase to 0.32 billion USD of FDI in 1988 but also meant a
greater number of projects were included. The important role of exports has been reflected via its
contribution to the national income, yielding from nearly 3% in 1986, gradually increasing to about
35% in 1997 and reaching 73% in 2012. For the period, the trade balance was mostly negative, the
difference between exports and imports was rather small.
Figure 5 shows the evolution of Vietnamese trade from 1996 to 2012 including four important factors
of trade. We see that there was a limited increase in exports and imports until 1992 and the number of
importing countries grew slightly between 1986 and 1992. However, from 1993 to the end of the
period, there was a slightly increasing trend in the country’s exports and imports and its trade balance
was mostly negative until 2012 when exports exceeded imports.
32 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Figure 5. Evolution of Vietnamese trade and its country partners, 1986-2012
Source: Export, Import and number of destination markets data for 1986-2012 GSO
(website of GSO, Vietnam)
3.2.2. Period 1992 - 2003
The year 1992 was the first time that Vietnam amended its Constitution since Doi Moi. It included the
guiding principle that the country pursues a policy of peace, friendship and expands international
relations and cooperation with all countries in the world, irrespective of their political and social
systems. The 1992 Vietnam's Constitution also acknowledged for the first time the right to create a
multi-sector competitive economy thus encouraging each economic sector to create its own value of
first phase of modernization and industrialization.
The period saw significant initiatives by Vietnam ranging from normalizing diplomatic relations to
establishing trade relations with some important partners either an individual country or foreign
economic blocs. Figure 6 shows the various stages in the development of Vietnam's trade relations
with key trading partners such as USA, Japan, China, the EU and ASEAN. In general, the open
economic policy approved by the State Government has had a positive effect on trade in particular on
the flows of exports, which make an important contribution to GDP, create employment, attract
advanced technology and encourage local production. When Vietnam became an official member of
ASEAN in 1995 and APEC in 1998 and normalized its diplomatic relations with the United States of
America in 1994, it was a milestone in its integration into the global market. In 1995, Vietnam signed
the Vietnam-EC framework cooperation agreement and in 2000, it signed a bilateral trade agreement
with the US triggering an active period of trade cooperation. The CEPT-AFTA reflects the tax cuts and
exemption by member countries, which facilitate and liberalize trade between members.
Moreover, the extension of ASEAN to ASEAN plus three including Japan, China and South Korea
was an opportunity for Vietnam to consolidate relationship with these large economies in the region.
Firstly, China shares a common northern border with Vietnam and is a major supplier of competitively
priced raw materials for production and Vietnam has much in common with China culturally and in its
pattern of consumption.
0
50
100
150
200
250
-400000
-200000
0
200000
400000
600000
19
87
19
88
19
89
19
90
19
91
19
92
19
93
19
94
19
95
19
96
19
97
19
98
19
99
20
00
20
01
20
02
20
03
20
04
20
05
20
06
20
07
20
08
20
09
20
10
20
11
20
12
Exports (real million USD) Imports (real million USD)
Balance (real million USD) Number of importing country partner
33 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Figure 6. Historical trade foundation of Vietnam and its country group partner
Source: Author's research
Secondly, Japan, which is a leading technological power, is a major investor and aid partner for
Vietnam. South Korea has significantly contributed to Vietnam's economy, on the one hand emerging
as Vietnam's number one investor, with $23.5 billion FDI in more than 3,000 projects (Park, 2012) and
on the other hand, being one of the top five largest customers in Asia for Vietnam's export products.
Vietnam also succeeded in exporting annually to South Korea an average of 2.1 million USD in the
VN-EU
1992
1995
1997
Signed textile agreement
Signed Vietnam-EC framework cooperation agreement
Vietnam joined ASEAN-EU cooperation agreement
VN-ASEAN
1995
1996
2002
1997
Signed the agreement on the Common Effective Preferential
Tariff Scheme for ASEAN Free Trade Area
Vietnam implemented first package for the CEPT scheme
ASEAN plus three including China, Japan and South Korea
ASEAN-Japan Free Trade Agreement
ASEAN-China Free Trade Agreement
1998 Became official member of APEC VN-APEC
VN-USA 1995
2000
Formal normalization of US-VN diplomatic relations
US-VN bilateral trade agreement (BTA) was signed
34 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
1986-1991 period and the annual export value exceeded 5 million USD on average in the second
period, indicating strong progress in trade tie between the two countries.
Figure 7. Exports by country and country group (%)
Source: Author's diagram based on GSO data
Figure 7 compares the export performance of Vietnam in some key markets. From the diagram, it is
clear that Vietnam's exports to those markets increased in all three periods.
It is noteworthy that exports to the ASEAN, EU, USA, Japan and China markets accounted for larger
share of Vietnam's total exports in the second period after Doi Moi showing how much progress
Vietnam achieved through trade liberalization. In fact, the total share of exports by value to those
markets remarkably doubled in the second and third period compared with first period after Doi Moi.
While ASEAN and Japan accounted for a largest share of Vietnam's total exports in the first period,
which was maintained during in the second and third one, the EU and the USA share rose spectacularly in
the second and third period. In particular, we should note that China has become a major export market for
Vietnam in the second and third period following its economic reform.
ASEAN
13%
EU
4.93%
Japan
16.08%
China
0.29% US
0%
The Rest
66%
Period 1986-1991
ASEAN
18.81%
EU
20.99%
Japan
18.71% China
7%
US
6.06%
The Rest
28.03%
Period 1992-2002
ASEAN
15.99%
EU,
17.1%
Japan, 12.02%
China, 10.8%
USA,
16.68%
The Rest
27%
Period 2003-2012
35 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Trade liberalization under the framework of the agreement on CEPT3 grants preferential tariff rates of
0-5% to member countries of the ASEAN. This policy enabled Vietnamese products to gain access to
a large regional market before expanding into the global market. Vietnam today has fully implemented
its commitment to cut its preferential tariff rate to 0-5%. In fact, the absolute value of Vietnam's
exports to ASEAN countries increased in line with the preferential tariff elimination roadmap
confirming its active participation in regional and international markets.
3.2.3. Period 2003- 2012
The third period is characterized by Vietnam's close integration into the world economy, as the country
became a full member of the WTO in 2006. Membership of the global trade community provides
Vietnam with support in seeking more opportunities for its firms to find export and import partners, for
it to increase trade revenue, and attract advanced technology from large intercontinental corporations.
However, just as many other members of the WTO, Vietnam has to comply with the general principles
of the multilateral trading system namely MFN4, transparency, and protection for domestic industry
through tariffs only and continued liberalization.
In particular, Vietnam is currently involved in comprehensive cooperation with some regional
economic associations. Firstly, the AEC established by leaders from the ASEAN bloc of nations, will
soon become effective at the end of 2015. As a single market, the AEC will help promote free flows of
goods, services, investment and employment in the bloc by eliminating tax on most goods and services
between ASEAN countries. Vietnam in turn, has committed itself to eliminate import taxes for about
80% of tariff lines as well as to remove of all tariff lines by 2018. Secondly, in 2009, Vietnam started
the first round of negotiations for the TPP agreement5. The TPP agreement is considered to be one of
the most important trade negotiations for Vietnam because it would give the country favorable
conditions to expand the export market and indirectly improve the quality of employment. Thirdly,
Vietnam would benefit from the upcoming free trade agreement with the EU expected to be concluded
in late 2015. After starting negotiation the free trade agreement with the EU in 2012, Vietnam has
actively initiated many free trade talks with EU countries and expects to achieve its goals of improving
the environment, competition, sustainable development and free investment.
3 CEPT reduction schedule affects four categories of product/items summarized in the Inclusion List (IL), the Temporary
Exclusion List (TEL), the Sensitive List (SL) and the General Exception List (GEL). Vietnam announced the classification of
goods in four lists including 4232 items are listed under IL, 1800 items under TEL, 51 items under SL, 202 items under GEL.
Major goods on Vietnam's TEL have been automobiles with more than 16 seats, household appliances, bicycles, toys,
cosmetics, textiles/garments, iron and steel products. The SL includes 26 items unprocessed agricultural products such as
livestock, meat, fish product, eggs, brown rice and sugar that are protected by quantitative import restrictions. GEL covers
196 products and their importation is restricted for reasons of environmental protection, national security, public moral and
health. 4 MFN is the most important discipline of the WTO referring to the fact that trade benefits granted to one country must be
immediately and unconditionally extended to like products of other countries. The national treatment, that complements the
MFN principle, prohibits a member from discriminating between imported products/services and domestically produced
products and services. 5 TPP is a multilateral free trade negotiation aiming to integrate the economies of the Asia-Pacific region. At present, 11
countries taking part in TPP negotiations include New Zealand, Brunei, Chile, Singapore, Australia, Peru, United States,
Malaysia, Vietnam, Canada and Mexico.
36 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
In 2005, Vietnam approved a wide code of law thus demonstrating its progress in acknowledging these
general principles and bringing its law into compliance with them. Moreover, the government of
Vietnam increasingly through new legislation provided a greater stimulus to foreign investment. The
three new laws, law on investment, law on enterprises and commercial law represent Vietnam's legal
commitments beyond the mandatory WTO provisions on the privatization of state-owned enterprises
and the transparency of trade policy. In particular, the 2005 law on investment included both domestic
and foreign investors as a legally regulated category whereas the 1987 law on foreign investment
applied solely to the foreign investors. Specifically, the 2005 law was the first to mention guaranteed
protection of intellectual property rights for investors replacing the narrower notion of industrial
property rights in the 1996 law on foreign investment.
As mentioned, the IPR is considered to be a challenge for developing countries such as Vietnam
following toward the process of global economic integration. Improving the IPR system, on the one
hand presents Vietnam’s commitment in the WTO, on the other hand, assures the rights to both foreign
and local investors over the creations of mind. Moreover, building an IPR system may help Vietnam to
improve its FDI environment through FDI policies, the cost of patenting and the quality of examination
processes and therefore to encourage FDI projects. As Vu (2012) showed a positive and significant
correlation between Resident Patent and FDI suggesting that FDI contributed to the growth of
technological capability and stimulated the demand for patents in the fast growing ASEAN countries.
According to WEF (2010), among fast developing ASEAN countries, Vietnam was ranked as 109th out
of 139 countries for its IP protection. The country experienced a very low IPR enforcement index at
0.4 point. Although Vietnam has made remarkable progresses in attracting more FDIs, the number of
patents filed per million capita in 2008 was about the same as number of patents per million capita of
Thailand filed in 1991. The filing cost of patent is rather expensive at approximately $587 especially if
patentees have more claims in their patent applications. Regarding the quality of examination process,
Vietnam has a low average in patent granting rate with only 19% of resident patent filings that were
granted against 21% of non-resident patent filings.
Vietnam does not only play an active role in the world economy but also demonstrates its high
capability to confront potential risks from large trading partners. Although GDP increased steadily over
the three periods peaking in 2012, GDP growth rate fluctuated strongly. There was a particularly sharp
decrease in the year 2000 partly due to contagion of the Asian financial crisis in 1997. Vietnam's GDP
growth rate fluctuated within a 5-6% range in the third period but fell steeply in 2009 one year after the
global financial crisis in 2008. However, there was an upward trend for registered FDI, which reached
a peak of 71.72 billion USD in 2008 before declining slightly in the following years.
37 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Figure 8. Vietnam's economic indicators
Source: GDP and GDP growth rate, World Bank (WB)
FDI, General Statistics Office of Vietnam (GSO)
3.3. Export trade pattern
3.3.1. Composition of export commodities
3.3.1.1. General composition of commodities
During the period of the country's regional division, Vietnam's economy mainly depended on three
layers comprising a bottom layer based on rice growing, a middle layer dominated by mining in the
north and rubber plantations in the south and the third layer in the north created by the large-scale aid
packages from the Soviet Union and China. After the country's reunification in 1975 and before Doi
Moi, Vietnam's exports were limited to agriculture, forestry and fishery products in the form of raw
material and products having undergone preliminary processing. However, Vietnamese exports
prospered for several years following the adoption of the Doi Moi policy by the Government. For the
1986-1991 period, there were 5 commodity groups6 which were the main exports according to State
Plan and which accounted for almost all the annual export revenue of the country. Significantly, all
these commodity groups are primary and/or labor intensive showing that Vietnam possesses a
comparative export cost advantage.
Figure 9 compares the share of Vietnam's five export commodities as set out in the State Plans over the
1986-2012 period. Generally, the country tends to focus on the stimulation of natural resource and
agriculture-based exports. Even through, each export commodity group’s contribution may vary, it
remained vibrant. While the aggregate share of agriculture, handicraft and light industrial and fishery
products dominated Vietnam's exports early in the period, mineral, heavy industry, handicraft and light
industry products accounted for a larger share of the country's exports later in the period. Over this
period, handicraft and light industry products experienced the greatest increase in export value whereas
mineral and heavy industry products saw a sharp increase in their export value in the third year after
Doi Moi then leveled off and fluctuated slightly until 2012.
6 Five commodity groups include mineral and heavy industrial products, handicraft and light industrial products, agriculture
products, forestry products and fishery products.
0
10
20
30
40
50
60
0
1000000
2000000
3000000
4000000
5000000
6000000
7000000
19
87
1
98
8
19
89
1
99
0
19
91
1
99
2
19
93
1
99
4
19
95
1
99
6
19
97
1
99
8
19
99
2
00
0
20
01
2
00
2
20
03
2
00
4
20
05
2
00
6
20
07
2
00
8
20
09
2
01
0
20
11
2
01
2
Per
cen
tag
e, %
Mil
lio
ns
US
D
Real GDP GDP growth rate Real registered FDI
38 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Figure 9. Share of export commodities, % of the total
Source: Author's diagram based on GSO's data
Mineral and heavy industry products increased their export value by an extraordinary amount,
followed by handicraft and light industry products. However, forestry products experienced the least
average increase. (see Table 6).
However, within the mineral and heavy industry product group, crude oil exports were responsible for
57% on average of the total after 1989. In 1991, 1992 and 1993, crude oil was the dominant export in
the group reaching 83%, 84% and 83% respectively.
The contribution of crude oil only became less important after 2010, falling to 22%. Crude oil is a
specific industry operated by the Vietnam National Oil and Gas Group, the sole company authorized
by the Government to explore, produce and export oil to the world market.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
19
86
19
88
19
90
19
92
19
94
19
96
19
98
20
00
20
02
20
04
20
06
20
08
20
10
20
12
Other products
Fishery products
Forestry products
Agriculture products
Handicraft and Light
industrial products
Mineral and heavy
industrial products
39 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Table 6. Commodity composition of Vietnam's exports (1996-2000, millions USD and %)
Year
Mineral and
heavy
industrial
groups
Handicraft &
Light industrial
products
Agriculture
products
Forestry
products
Fishery
products
1986
63.4
8
227.5
28.8
318.6
40.4
71.6
9.1
106
13.4
1988
66.9
6.4
383.8
37
349.2
33.6
59.2
5.7
178
17.1
1990
617
25.7
635.8
26.4
783.2
32.6
126.5
5.3
239.1
9.9
1992
954.8
37
349.5
13.5
827.6
32.1
140.8
5.5
307.7
11.9
1994
1,167.6
28.8
938.2
23.1
1,280.2
31.6
2.8
11.6
556.3
13.7
1996
2,085
28.7
2,101
29
2,159.6
29.8
212.2
2.9
696.5
9.6
1998
2,609
27.9
3,427.5
36.6
2,274.3
24.3
191.4
2
858
9.2
2000
5,382.1
37.2
4,903.1
33.9
2,563.4
17.7
155.7
1.1
1,478.5
10.2
2002
5,304.3
31.8
6,785.7
40.6
2,396.6
14.3
197.7
1.2
2,021.7
12.1
2004
8,633
32.6
10,920.2
41.2
3,383.6
12.8
180.5
0.7
2,401.2
9.1
2006
14,428.6
36.2
16,389.5
41.2
5,352.4
13.4
297.5
0.8
3,357.9
8.4
2008
23,209.3
37
24,896.3
39.7
9,239.5
14.7
468.7
0.7
4,510.1
7.2
2010
22,402.8
31
33,336.9
46.1
10,639.4
14.7
803.9
1.1
5,016.9
6.9
2012
48,228.1
42.1
43,298.6
37.8
15,463.4
13.5
1363.7
1.2
6,088.5
5.3
Source: GSO
3.3.1.2. Major export products, regions and economic groups
Apart from the exception of crude oil, the other major export products of Vietnam can be freely traded
subject to Vietnam's legislation. There are also numerous firms active in other export sectors. As
previously analyzed, Vietnam is strong in manufacturing and exporting resource intensive and
agricultural products although they create low value in the global value chain.
- Major exports
Major exports are limited to the handicraft, light industry and agriculture commodity group. This
section analyzes sub-commodities from the above-mentioned commodity groups according to SITC17
classification (Revision 3). Although the relative contribution of these sub-commodities in Table 7
have changed since Doi Moi, the major exports continued to be labor-intensive, confirming the
7 SITC was first issued by the United Nations' Secretariat in 1950. This is a classification of goods used to classify
the exports and imports of a country to enable comparisons between different countries and years.
40 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
dominance of traditional export products in total exports and the absence of structural changes in the
major exports during the period.
Articles of apparel and clothing accessories are the largest sub-commodity making significant
contribution to Vietnam's total exports. Table 7 shows 18 major export products for four five-year
periods since Vietnam's economic reforms in 1986. The respected share of each product in total
exports changed over these periods. In the first period, fishery products exports earned an average of
861 million USD, followed by articles of apparel and clothing accessories with an average of 675
million USD and rice products ranking third in export value for these periods. The ranking for these
products remained unchanged until the third period when footwear products emerged as the second
largest export sector, achieving an average of 5,398 million USD and remained in second position for
the 2000-2005 period. During the same period, articles of apparel and clothing accessories reached the
strong first ranking in total product exports.
All major product exports clearly increased throughout the periods with the exception of the wool,
rubber, meat and meat preparations. In the 2000-2005 period, exports of these products suddenly fell to
zero indicating that those product sectors had lost their market in the five years preceding Vietnam's
WTO membership. Significantly, the average export value of footwear products showed an
extraordinary 220-fold increase in the fourth period after Doi Moi compared with the first. The export
revenue from articles of apparel and clothing accessories rose steeply, 126 fold compared with the first
period. The value of tin and cinnamon product exports increased slightly by 0.24% and 0.06%
respectively in the fourth period compared to the first.
There was a structural change after 2006 for some major export products (see Table 8). That year also
marked by the important turning point of Vietnam's membership of the WTO. On 7 November 2006,
the General Council of the WTO formally approved Vietnam's accession package and the country
became a full member of this organization one year later. During the period, export revenue generally
increased and there was a greater diversification of export products At this time, the number of major
export products doubled to 38 compared to the situation before membership. Among the major export
products, the value of exports of articles of apparel and clothing accessories, footwear, fishery
products, wood and wood products, rice, rubber and coffee grew significantly except for 2009 when
the country was severely affected by the global financial crisis, which caused an economic downturn.
However, exports fell for the two product categories: bicycles and parts of bicycles, groundnuts
shelled, which before Doi Moi had been mainly exported to the Soviet Union and Eastern bloc
countries. By 2012, embroidery products and art articles of pumice lacquer were no longer amongst
Vietnam's 38 largest exports. Vietnamese exports of phones of all kinds and their parts as well as
cassava and cassava products have grown impressively, reaching 12,746.5 million USD and 1,351.4
million USD respectively.
Traditionally as mentioned, the structure of Vietnam' exports was mainly linked to agriculture, fishery
and forestry items. However, since 2006, light industrial products have begun to perform better on
export markets with significant contribution of several sectors: electronic parts, computer and their
41 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Table 7. Major exports in each five year period (millions USD)
Major exports 1986-1990 1991-1995 1996-2000 2001-2005
1. Articles of apparel and clothing accessories 675 1,912 7,744 17,541
2. Fishery products 861 2,193 4,788 11,191
3. Footwear 52 511 5,398 11,454
4. Rice 661 1,969 4,438 4,427
5. Coffee 343 1,207 2,598 2,594
6. Articles of vegetable fibers 175 12 205 687
7. Articles of rattan, bamboo 194 92 83 0
8. Articles of art 73 81 135 0
9. Embroidery 144 76 143 0
10. Jute rugs 56 7 32 0
11. Wool rugs 53 38 4 0
12. Rubber 207 514 885 2,216
13. Coal 117 319 517 1,480
14. Vegetables and fruits 272 166 534 1,131
15. Pepper 60 113 462 608
16. Cashew nuts, shelled 40 270 603 1,576
17. Tin 41 93 72 51
18. Meat and meat preparations 77 121 88 0
Source:GSO
42 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Table 8. Major exports in period 2006-2012 (millions USD)
Major exports 2006 2007 2008 2009 2010 2011 2012
1. Articles of apparel and clothing accessories 5,854.8 7,732 9,120.4 9,065.6 11,209.8 13,211.7 14,416.1
2. Footwear 3,596.9 3,999.4 4,769.8 4,071.2 5,123.3 6,549.4 7,263.9
3. Fishery products 3,357.9 3,763.4 4,510.1 4,255.3 5,016.9 6,112.4 6,088.5
4. Electronic parts computer and their parts 1,807.8 2,165.1 2,640.3 2,763 3,590.1 4,662.1 7,848.8
5. Wood and articles of wood 1,943 2,384.6 2,767.1 2,597.6 3,444.5 3,961.1 4,665.5
6. Rubber 1,286.3 1,393.8 1,604.1 1,227.1 2,386.1 3,234.4 2,860.1
7. Rice 1,275.8 1,490.1 2,895.9 2,666 3,249.5 3,659 3,673.6
8. Coffee 1,217.1 1,916.6 2,113.7 1,730.5 1,851.4 2,760.2 3,674.4
9. Coal 914.8 999.7 1,388.4 1,316.5 1,614.5 1,632.1 1,239.8
10. Electrical wire and cable 705.7 882.3 1,008.9 891.7 1,315.9 443.5 618.7
11. Cashew nuts, shelled 503.8 645.1 915.8 849.6 1,136.9 1,473.1 1,470.1
12. Rucksacks, bag, pockets, wallets 502 627.1 773 824.1 985.4 1,285.3 1,522.5
13. Plastic produce 452.3 709.4 933.6 867.3 1,130.1 1,373.7 1,595.5
14. Iron and steel products 389.3 403.2 706.2 620.2 832.9 1,130 1,386
15. Ceramic articles 274.4 334.8 344.3 267.1 317 359.2 440.4
16. Vegetables and fruit 259 305.6 406.4 438.8 460.2 622.5 827
17. Articles of rattan, bamboo, rush and leaf 203.6 233.5 199.5 1.5 5.9 181.5 185.9
18. Carpets 10.4 13.2 24.1 20.4 21.4 21.2 25.1
19. Pepper 186.5 271.4 311.4 348.3 421.4 732.4 793.6
20. Preparations of cereals, flour, milk 151.2 194 258.5 267.3 385.1 377.4 410.7
21. Phones all of kinds and their parts 0 0 0 0 0 6,396.7 12,746.5
22. Art articles of pumice lacquer 119.5 217.8 385.4 1,296.1 14.1 0 0
23. Bicycles and parts of bicycles 110.5 81.1 89.1 85.8 93.6 44.6 18
24. Tea 110.4 133.4 147.3 180.2 200.5 205.5 224.8
43 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
25. Cassava & Cassava products 0 0 0 0 0 0 1,351.4
26. Embroidery products 98.1 111.8 110.5 129.2 153.9 3.6 0
27. Diary produce 90.1 16.3 29.6 23.6 33.6 66.5 122.7
28. Toys 32.8 35 110.8 118 155.7 193.7 212.3
29. Meat and meat preparation 26.3 48.4 58.9 45 40 58.9 68
30. Tin 20.4 28.5 41.5 9.06 27.5 59.1 35
31. Vegetable oils and fats 15.3 49.3 99.5 77.3 98 209.5 314.4
32. Cinnamon 14.3 16.2 5.2 5.7 5.7 5.3 5.7
33. Ground nuts, shelled 10.4 31.2 13.7 21.5 22.5 7.1 5.6
34. Chromites 2,7 0 0 0 0 0 0
35. Sugar 2.3 4.7 5.01 1.5 0.8 173.8 47.1
36. Articles of precious stones and metal 2.08 3.7 415.3 2,730.7 2,823.9 2,668.7 545.8
37. Maize 0.84 1.2 6.2 1.2 1.1 2.01 7.3
38. Electrical Energy 0,027 0.28 0.74 11.06 56.6 79.3 28.9
Source: GSO
44 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
parts, phone all of kinds and their parts. Vietnam has become an important exporter of electricity and its
revenue grew impressively by 37% in 2011 compared with 2006.
- Exports by geographical region
Since the economic reforms in 1986 and the official Government' directive to open up its market,
Vietnam has been prepared to be a trade partner for all countries of the global market and export value
as well as the number of trading partners have continuously increased. In fact, Vietnam has been able to
supply almost all countries across the continents of Asia, Europe, America, Africa and Oceania. Figure
10 illustrates by all geographical regions the rising trend in the country's exports where Asia is the
largest export market, followed by Europe and America meanwhile Oceania imports least from
Vietnam. However, the European and American markets swabbed position after 2004 when export
value of the American markets exceeded that of European markets. Initially, from 1986 to 1993, the
value of the country's exports to the European market was always higher than to Americas. Specifically,
in 1986, the value of Vietnam's exports to Europe was 446.9 million USD while that to Americas was
only 14.2 million USD. By 2003 and 2004, the value has almost converged and by 2005, America has
caught up, overtaken Europe and has become as the second largest importing market for Vietnam after
Asia.
Figure 10. Exports by geographical region
Source: GSO export database
Most of Vietnam' export markets were dynamic with the exception of Africa and Oceania where
although the value of Vietnam's exports rose slightly in 2006 then fell back and stagnated between 2010
and 2012. Meanwhile, Vietnam's export growth on European and American markets was accelerating,
particularly after 2008. Surprisingly, in the two years immediately following Doi Moi, Asia - Vietnam's
largest export market performed less well than Europe. However, after 1988, this market became
dominant and Vietnam's most dynamic import market. After falling to 26,648 million USD in 2008, the
value of Vietnam's exports to the Asian market increased impressively achieving by 61,368.3 million
USD in 2012.
0
20000
40000
60000
80000
100000
120000
Mil
lio
ns
US
D Ocenia
Africa
America
Europe
Asia
45 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
- Exports by country group
The creation of economic groups aim to facilitate trade between member countries through tariff or non-
tariff measures. There are many regional economic groups based on shared characteristics of geography,
politics or economics, in addition to the WTO largest trade organization. Figure 11 shows the value of
Vietnam's exports to following country groups: ASEAN8, APEC
9, EU
10 and OPEC
11 the first two of
which Vietnam is a full member.
Among these groups, APEC emerges as the biggest and the most dynamic export market of Vietnam.
From the beginning of Doi Moi until today, APEC has always taken the lead in attracting export
products from Vietnam. In 2012, Vietnam’s revenue from exports to APEC hit its highest level in 1986,
78,028.2 million USD. The EU was the second largest import market for Vietnam, followed by ASEAN
with OPEC being the least dynamic market for Vietnam's exports over this period.
Figure 11. Exports by country group
Source: GSO export database
Although, Vietnam's exports to APEC were very strong, they fell back to 38,802.1 million USD in 2009
due to the effect of the global financial crisis, before recovering and reaching a new record level in 2010.
Meanwhile, exports to the EU and ASEAN markets were steadily increasing and almost converged in
1999.
8 ASEAN is a political and economic organization, which includes Indonesia, Malaysia, the Philippines, Singapore, Thailand,
Brunei, Cambodia, Laos, Myanmar and Vietnam.
9 APEC seeks to promote free trade and economic cooperation throughout the Asia-Pacific economies. Its members are
Australia, Brunei Darussalam, Canada, Indonesia, Japan, South Korea, Malaysia, New Zealand, Philippines, Singapore,
Thailand, United States, Chinese Taipei, Hong Kong, People' Republic of China, Mexico, Papua New Guinea, Chile, Peru,
Russia and Vietnam.
10
EU covers the following 28 sovereign states Austria, Belgium, Bulgaria, Croatia, Cyprus, Czech Republic, Denmark,
Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands,
Poland, Portugal, Romania, Slovakia, Slovenia, Spain, Sweden and United Kingdom.
11
OPEC covers 12 countries, which are Qatar, Libya, the United Arab Emirates, Algeria, Nigeria, Angola, Iran, Iraq, Saudi
Arabia, Kuwait, Ecuador and Venezuela.
0 20000 40000 60000 80000 100000 120000
1986
1989
1992
1995
1998
2001
2004
2007
2010
Millions USD
ASEAN
APEC
EU
OPEC
46 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
3.3.2. Export composition commodity by destination market
3.3.2.1. Sectoral dimension and country groups
Immediately after the introduction of Doi Moi, the country mainly produced raw materials. Just as many
other developing countries in the region, it focused on exploiting its strength which are natural resources
and competitive labor cost. In the short-term when Vietnam had just moved past the stage of shortage of
goods, hyperinflation, capital scarcity and skilled labor paucity, concentrating on primary products was
inevitable. Figure 12 illustrates the proportion of two categories over the two consecutive periods of
1986-1996 and 1997-2012.
In the first period after Doi Moi, the export value of primary and manufactured products gradually
increased. For primary products, the increase was relatively steady however, exports of manufactured
products fluctuated somewhat. From 1986 to 1996, exports of primary products outperformed
manufactured products. However, since 1997, the export value of manufactured products increased
significantly and exceeded that of primary products reaching 46.2 billion USD in 2008 before falling
back sharply to 43.8 billion USD in the following year due to the effect of global financial crisis. For the
same years, however, the value of exports of primary products fell by 6.9 billion USD. More
significantly, from 1997 to 2002, the value of exports of manufactured products has caught up with
primary products and was constantly higher since 2003. In 2012 in particular, the value of exports of
manufactured goods reached 100.4 billion USD, three times as much as primary products.
These values marked the outstanding achievement of Vietnam in changing the structure of export
products and enhancing its comparative advantage by exporting products with high skilled labor content
as well as its success in finding new raw material inputs to replace natural resources. We note that, the
value of exports of primary and manufactured products varies depending on the country group
destination.
Throughout this period, APEC continues to be Vietnam's largest export market for primary and
manufactured products followed by ASEAN and the EU. However, export level to APEC varied over
the 2004-2012 period during which the value of exports of manufactured products overtook primary
products.
47 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Figure 12. Export value of primary and manufactured products (millions USD)
Figure 13. Export value by country groups (millions USD)
For primary products
For manufactured products
Source: Author' diagrams based on GSO data
We can see that APEC has imported manufactured products since 2007 reaching approximately 52
billion USD in 2012, which is twice the value for imports of primary products. In contrast, the value of
primary product exports from Vietnam to ASEAN tends to be higher than for manufactured products
whereas the value of manufactured product exports to the EU exceeded the value of primary products
for the whole period and contributed 16.7 billion USD to Vietnam's export revenue in 2012.
Figure 14 compares the value of Vietnam's exports of primary and manufactured products categories by
country groups from 2004 to 2012. In general, the trend line broken down for the different categories
0
1000
2000
3000
4000
5000
6000
7000
8000
19
86
19
87
19
88
19
89
19
90
19
91
19
92
19
93
19
94
19
95
19
96
Primary products
Manufactured products
Others
0
20000
40000
60000
80000
100000
120000
140000
160000
1997 1999 2001 2003 2005 2007 2009 2011
0
5000
10000
15000
20000
25000
30000
35000
40000
45000
ASEAN APEC EU Others
0
20000
40000
60000
80000
100000
120000
ASEAN APEC EU Others
48 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Figure 14. Export value of primary products sector
Source: Author's diagrams based on GSO data
differs between the ASEAN, APEC, EU and others country groups. Primary products are commodities
classified at the 1-digit level under the SITC: food and live animals, beverages and tobacco, crude
materials, inedible except fuels, animal and vegetable oil, fats and waxes. Manufactured products
comprise chemicals and related products, manufactured goods classified chiefly by materials,
machinery, transport equipment and miscellaneous manufactured articles.
Food and live animals, crude materials and inedible except fuels dominate and account for the largest
proportion of the primary product sector, achieving a value of nearly 12 billion USD to the APEC
market in 2008. Beverages and tobacco, animal and vegetable oils, fats and waxes, however had a
smaller share of this market earning 0.5 billion USD and 0.3 billion USD in 2012 respectively in export
revenue. Especially, APEC is a biggest importing country group in all categories of primary sector
through the years. For this specific category of food and live animals, APEC is the largest country group
importing market and pioneering role compared with the other country groups. The ASEAN country
0
2000
4000
6000
8000
10000
12000
14000 M
illi
on
s U
SD
Figure 14a. Food and live animals
ASEAN
APEC
EU
Others
0
50
100
150
200
250
300
Mil
lio
ns
US
D
Figure 14b. Beverage and tobacco
ASEAN
APEC
EU
Others
0
2000
4000
6000
8000
10000
12000
14000
Mil
lio
ns
US
D
Figure 14c. Crude material, inedible
except fuels
ASEAN
APEC
EU
Others
0
50
100
150
200
250 2
00
4
20
06
20
08
20
10
20
12
Mil
lio
ns
US
D
Figure 14d. Animal and vegetable
oils, fats and waxes
ASEAN
APEC
EU
Others
49 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
group was the least dynamic for this category, after the EU. For the beverage and tobacco category,
APEC has become a major import market with levels rising rapidly since 2009, followed by ASEAN
with the EU coming third. Vietnam's revenue from exports to APEC of crude materials, inedible except
fuels strongly fluctuated whereas the value of exports to ASEAN only fluctuated slightly. However,
there has been strong rising trend in the value of exports to the APEC and ASEAN markets of the
animal, vegetable oils, fats and waxes categories.
Figure 15. Export value of manufactured products sector
Source: Author's diagrams based on GSO data
As Figure 15 shows, there is a generally increasing trend in the value of exports to all country groups in
the four categories of the manufactured products sector in almost all cases. APEC remains the leading
country group importing all kinds of manufactured products from Vietnam while the role of ASEAN
and EU in this regard is broadly equivalent. ASEAN imports more chemical and related products and
0
500
1000
1500
2000
2500
3000
Mil
lio
ns
US
D
Figure 15a. Chemical and related
products
ASEAN
APEC
EU
Others
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
Mil
lio
ns
US
D
Figure 15b. Manufactured goods
classified chiefly by materials
ASEAN
APEC
EU
Others
0
5000
10000
15000
20000
25000
Mil
lio
ns
US
D
Figure 15c. Miscellaneous
manufactured articles
ASEAN
APEC
EU
Others
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000
Mil
lio
ns
US
D
Figure 15d. Machinery and transport
equipment
ASEAN
APEC
EU
Others
50 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
manufactured goods classified by materials than the EU but imports less miscellaneous manufactured
articles and machinery and transport equipment than the EU. Although APEC is the largest export
market for all four categories, the chemical and related products category represents the smallest share
and miscellaneous manufactured articles account for the largest share, earning nearly 20 billion USD in
2011.
3.3.2.2. Market and product concentration During the process of political reform and international economic integration, Vietnam has faced many
problems arising from strong competition and market vulnerability. Through the Herfindahl-Hirschman
market and product concentration index, it is possible to measure the dispersion of trade value across an
exporter's partners and products. Specifically, a higher index indicates that one country concentrates on
a very few markets or products. In other words, a country with higher indexes tends to be more
vulnerable than other countries. In the case of Vietnam, we use the HH product concentration index
value derived from the UN Comtrade data for all export markets and products at the 6-digit HS codes
level as issued in 1996.
Figure 16 shows the fluctuations of the HH market and product concentration index for the 2000-2012
period. The two trend lines clearly are falling which is a positive signal for Vietnam's exports since a
lower index directly reflects a higher market and product diversification thereby diminishing the threat
of vulnerability. However, after 2006, the HH product concentration index decreased faster than the HH
market concentration index coinciding with a growing number of export products whereas the number
of Vietnam's export markets has gradually decreased after 2005 having peaked at 218 markets in 2002.
This positive signal shows that the country has mainly concentrated on diversifying export products
across markets rather than seeking more markets.
Figure 16. Vietnam's H-H market and product concentration index
Source: Author's diagram based on UN COMTRADE data
Figure 17 compares the HH market and product indices of several ASEAN countries for aggregate
markets and products in 2012. Using the HH market concentration index, Thailand emerges as the least
vulnerable country since its export markets and products are the most highly diversified, followed by
Singapore and Vietnam ranking third. However, Cambodia and the Philippines are the most vulnerable
0
0.02
0.04
0.06
0.08
0.1
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
HH Market Index HH Product Index
51 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
for export market volatility. As shown in Figure 17, Singapore has a relatively low diversification of
export products with Cambodia having the lowest whereas, Vietnam and Thailand have the highest
diversification of export products.
Figure 17. ASEAN countries' H-H market and product concentration index in 2012
Source: Author's diagram based on UN COMTRADE data
3.4. Conclusion
In the nearly 30 years since Vietnam implemented its Doi Moi policy, the national economy has
witnessed a breakthrough in its economic development and trade growth. From a closed economy with
a limited number of trading partners in Eastern Europe and the former Soviet Union, Vietnam has been
transformed into an open economy, actively engaging in the global market. The country so far has also
been able to diversify to a maximum of 3610 export products at the 6-digit level and has become a
relatively dynamic exporting country in the region.
Thanks to Doi Moi since then, the fundamental economic indicators GDP, FDI and exports have
continued to improve. Inflation in particular has fallen back to a 2-digit rate from a 3-digit hyperinflation
rate in 1986. More importantly, along with the growth in exports, Vietnam has managed to attract more
capital investment and advanced technology from multinational corporations in developed countries.
Although, the country experienced certain negative effects caused by the regional financial crisis in
1997 and the 2008 global financial crisis resulting in a sharp drop in exports to 57 billion USD in 2009
compared with 62.6 billion USD in 2008. However, the total value of exports then soon recovered and
peaked in 2012 at nearly 115 billion USD.
With its strategy of economic integration including the open trade policy, Vietnam has assumed an
important role in the regional and global economy including ASEAN and the WTO. This strategy has
enabled the country to develop additional markets and increased its export volumes. The normalization
of diplomatic relation with the US represented an important launching pad for the country enabling it to
sign trade agreements and establish a number of trade links with USA's allies.
Moreover, another striking feature was the transformation of Vietnam' export product structure after Doi
Moi. Previously, the country had used its comparative advantage in exploiting natural resources, raw
0.0617
0.1297
0.0677 0.0665
0.0934
0.0609 0.0481
0.0195
0.1078
0.0273 0.03 0.0238
0.0757
0.0117
Vietnam Cambodia Indonesia Malaysia Philippines Singapore Thailand
HH Market Index HH Product Index
52 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
materials and low-skilled labor. Indeed, the primary sector accounted for the larger share in total exports
until 2001. From 2001 onwards, the country focused on exporting manufactured products where
machinery and transport equipment, miscellaneous manufactured articles were dominant and by 2012,
the export revenue of this category was almost three times the revenue from primary products. Thanks
to its open trade policy and integration strategy, Vietnam has gradually moved out of poverty and
improved the quality and value of export products by reducing the share and low cost unskilled labor
products.
The following chapter concentrates on examining Vietnam's export pattern through the gravity model
where its two main factors among others include the national income of the importing country and the
geographical distance between Vietnam and its importing partners. It is important to note, the aggregate
export value at country level and its determinants are observed for several years, which in some cases
may cause an endogeneity problem, the presence of which needs to be checked.
53 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
CHAPTER 4. INTERNATIONAL EXPORT TRADE FLOWS OF VIETNAM: A
GRAVITY MODEL APPROACH
SUMMARY
This chapter attempts to define the export pattern of Vietnam with its trading partners by using the
gravity model, showing one-way exports of Vietnam to the world market between 1997 and 2009. Since
the Hausman test does not support the random effects model in which we assume that unobservable
effects within destination markets do not bias the outcome variables, the fixed effects model is
consequently preferable. The role of the economic size indicator reaffirms its high relevance to export
trade. The geographical distance gives a negative and insignificant sign suggesting that the trade barrier
needs to be examined in the context of Vietnam' s trading with the rest of the world. As such, we have
employed the MRT for which the exchange rate between Vietnamese currency and an importing
country’s currency acts as a good proxy. Although this indicator encourages Vietnamese exports, it is
found to be significant for the OLS regression model. ASEAN as a regional economic integration
organization plays a positive role and therefore constitutes a significant absorptive factor. Interestingly,
high labor productivity importing countries tend to have a lower demand for export products from
Vietnam and countries which share a common border with Vietnam also become important and positive
trading partners.
54 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
4.1. Introduction
In recent decades, globalization has led to a pattern of multilateral trading between countries or free
trade associations. Exporting countries have been able to gain major benefits particularly through
offering diverse products giving global consumers a wide choice. Such gains vanish where highly
protected economies impose constraints on production and consequently consumption. Albeit, larger
importing country markets lead to tougher competition between exporting countries who need to be able
to surmount typical trade barriers such as geographic, social and economic hurdles. Some regional
economic groups offer member countries preferential treatment for trade and stimulate production
capacity as well as technological development, which highlight the role of such an area in fostering free
trade.
After two decades open to the world’s economy, Vietnam has achieved remarkable growth on some
economic indicators such as GDP, Exports and Foreign Direct Investment, etc. Since the US embargo
was lifted in February 1994, Vietnam has become an active member of ASEAN. Vietnam also became
a member of APEC in 1998. Those steps paved the way for this small economy to stimulate bilateral
and multilateral trade hence improved exports particularly when it became a full member of the WTO in
2007. Between 1997 and 2009, exports and GDP gradually increased and exports peaked at over 50
billion USD in 2008, falling slightly to 50 billion USD in 2009. Meanwhile, GDP saw a gradual rise
over the same period, increasing from 100 billion USD to over 250 billion USD in 13 years. Overall, the
export-to-GDP ratio increased remarkably and the rate of GDP growth increased, reaching 0.25% in
2009 and dropping sharply to 0.2 % partly due to the effect of the world financial crisis.
Many empirical studies on bilateral trade have confirmed the success of the gravity model in explaining
the intra-trade among countries involving in a single free trade area such as the EU, APEC, CARICOM
or the inter-trade between two economic blocks. Serlenga and Shin (2004), Rault et al. (2007), Mitze
(2012) and Sapir (2001) found that liberalization within the EU tends to lower prices through increased
competition and the removal of barriers to the mobility of goods leads to an increase in trade flows
within the commodity. However, the role of the EU in the study by Bastos and Silva (2010) changing
from positive to negative and insignificant to Portuguese exports once export firm-fixed effect, product-
fixed effect and firm-product fixed effects were examined implies that productivity of firms frame the
export pattern.
In regard to the demand side of a partial equilibrium, the income of each economy measured by GDP,
and/or GNP shows that country would expense more if it has the ability to earn more. Further, literature
on gravity model demonstrates a higher preference to heterogeneous import products rather than to
homogenous goods. To express differently, Nguyen (2009) found that homogenous goods are less
responsive to changes in income than heterogeneous goods. Regardless of time and space, GDP
consistently enhances trade capacity between countries even if many papers tend to confirm endogeneity
for these indicators.
Barriers in international trade include geographical, social and economic distance. Geographical
distance is regarded as a transport cost. In most of the paper using aggregated data on trade, distance
55 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
decreases systematically with export and import value. Although distance is a time-invariant factor,
using it as a proxy is problematic because the underlying assumptions vary over time. To resolve the
difficulty of using it in a fixed effect model, Brun et al. (2005) defined the augmented transport cost
function which includes indexes for infrastructure, price of oil in a random effect model showing that
distance did not vanish and providing support for the distance estimates in a range [-1.3; -0.8]. Bastos
and Silva (2010) conversely demonstrated that distance for a firm’s exports does not matter and such
firms may be able to charge a higher price and/or sell in larger quantities to more distant market, which
strongly supports the idea that a firm’s productivity is a key factor determining its export pattern.
How does the gravity model explain export flows of Vietnam? If we make some new assumptions on
certain typical factors of the full gravity model, this then makes it possible to solve the puzzle as to
whether a regional economic association such as ASEAN remains relevant for Vietnamese exporters.
There is little existing literature on the gravity model for Vietnamese bilateral and multilateral trade. Do
(2006) has used data on Vietnam’s exports and imports to 23 European countries between 1993 and
2004 to analyze trade potential with individual importing countries. Nevertheless, the paper is limited to
fixed and random effect test where an endogeneity problem may arise. Nguyen (2009) examined the
intra-trade between Vietnam and the other ASEAN countries under AFTA and applied a similar model
but the absence of some trading partners therefore in both cases meant that the conclusions about the
determinants were somewhat incorrect. In this chapter, we attempted to examine the pattern of
Vietnam’s exports to all its importing countries where in particular ASEAN countries’ trade with
Vietnam was examined. One important point to note is that geographical distance is negative but no
longer significant although it had been previously considered a good proxy for the transport costs of
international trade. This finding raises a question as to whether another efficient cost indicator such as
the multilateral resistance term should be substituted. The main findings also reveal that high-income
trading partners are rightly found to be important to Vietnamese exports, whereas high labor
productivity markets are not always a suitable target. The greater the difference between the GDP per
capita of the potential importing country and Vietnam, the lower the export value of Vietnam to such a
country. Moreover, the economic integration area – ASEAN group, helps Vietnamese exports to
member countries.
The remainder of the chapter is structured as follows: Section 4.2 explains the theoretical background
of the gravity model, section 4.3 describes the data used in the empirical analysis and the methodology,
section 4.4 presents the empirical results of the gravity model for Vietnam's export country trade and
section 4.5 offers some concluding remarks.
4.2. Theoretical background on the gravity model
The main branch of new trade theory is the gravity model firstly introduced by Tinbergen (1962) which
considers the interaction in trade between country pairs. The baseline model is a modified version of
Isaac Newton’s law of gravitation to predict movement of commodities between countries and
continents taking into account the economic size of the two countries and their distance.
31 2
0ij i j ijE Y Y D
56 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Where ijE is the flow of trade between country i and country .j iY and jY are the country 'i s and
country 'j s GDP and ijD is the geographical distance among two countries’ capital. The above model
could be put in the linear form as follows:
0 1 2 3( ) ( ) ( ) ( )ij i j ijLog E log Y log Y log D
As Christie (2002) specified that trade flow between two countries is proportional to the product of each
country’s economic mass measured by GDP, each to the power of quantities to be determined, divided
by the distance between the countries respective “economic center of gravity” generally their capital,
raised to the power of another quantity to be determined. In some cases, a group of dummy variables
can be added to the traditional model comprising of the country’s being a member of trade agreement,
sharing the common border or speaking the same language, etc.
The gravity model is considered to have connections to key elements of the classical trade theory such
as Ricardian and Heckscher-Ohlin models. Primary Ricardian model assumed that labor is only input
and countries trade is based on the differences between labor productivity which uses different
“technologies”. Moreover, the opportunity cost theory applies to the case when a country with a
comparative advantage in producing a good uses its resources most efficiently when it produces that
good rather than producing other goods. According to Heckscher-Ohlin model, relative endowments of
the factors of production including land, labor and capital determine a country’s comparative advantage.
A country will export that commodity which uses intensively its abundant factors such as labor, and
land while importing a commodity, which uses intensively its scarce factors. Meanwhile, the gravity
model describes trade relationship between North-South countries by showing a range of low-cost of
labor and high-technology products flows which generate trade pattern between country pairs.
The theoretical framework of the gravity model later focused on the supply and demand side and how
consumers’ wishes could be satisfied as to the utility of imported and domestically produced
commodities. Highly critical of some previous papers, Linnemann (1966), Aitken (1973), Geraci and
Prewo (1977) asserted that the gravity model is a reduced form of a four – equation partial equilibrium
model of export supply and import demand. Specifically, Linnemann (1966) assumed that the potential
demand of a country for another countries’ goods depends on its income or GDP and population.
Although his empirical result, which were based on cross-section data of trade flows have given
expected signs of the GDP variables, in some cases, GDP variables are not significantly different from
unity. As such, the effect of importing countries’ GDP should be examined in the context of world
GDP. This approach is “loose’ because it does not explain the multiplicative functional form.
Goldstein and Khan (1978) examined the association of a country’s export volume with price
determination in a simultaneous equation between the supply and demand sides of exports. In the
57 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
demand side equation12
, weighted average of the real incomes of the country’s trading partners
considered as one basic gravity factor is expected to be positive whereas the relative price13
is expected
to be negative. The supply side equation14
includes relative price of exports15
and an index of the
productive capacity of the country. Because an equilibrium model does not analyze lags effect, the
adjustment of export volume and prices takes place within a one-quarter period. By normalizing the
supply equation for the price of exports and solving the two equations to obtain the reduced-forms, these
authors expected that the weight average of the real incomes of the country’s trading partners16
yields
the positive effect on both export volume and price of exports. Moreover, a disequilibrium model in
which exports are assumed to adjust to the difference between demand for exports in period t and
actual flows in the previous period (1log (log log )d
t t tX X X 17
. Again, the weight average of
the real incomes of the country trading partners at time t is expected to have a positive effect on export
volume and price of exports at time .t Amazingly, empirical results for both equilibrium and
disequilibrium models show the expected positive signs of income elasticities implying that high
income importing countries were attractive destinations in respect with export volume of the exporting
countries.
Anderson (1979) referred to the gravity model as a pure Cobb-Douglas expenditure system where each
country is assumed to be completely specialized in the production of its own good and no tariff or
transport costs exist. The consumption of good i in country j is accordingly identified as ij i jM bY
where ijM denotes the multiplication of the price of good i and its quantity, jY is income in country j
and ib denotes the fraction of income spent on the product of country i . If income is supposed to equal
sales implying that ( )i i jY b Y then / .ij i j jM YY Y However, this simplest form of “gravity
model” may result in simultaneous equation bias because of the presence of .iY A country in reality
imports a variety of goods with price differences thus determining different expenditure shares in a less
12
Demand side equation is specified as 0 1 2log log( / ) logi
d i iX a a PX PXW a YW
13
This variable is specified by dividing price of exports by weighted average of the export prices of the country’s trading
partners
14
Supply side equation is specified as *
0 1 2log log( / )s
t t tX PX P Y
15
This variable is specified by dividing price of exports by domestic price index
16
Eight countries are included in the sample as follows: Belgium, France, Germany, Italy, Japan, the Netherlands, the United
Kingdom and the United States.
17
Where is the coefficient of adjustment and is the first difference operator
58 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
restrictive preference form such as the CES18
. By examining only the level of expenditure on traded
goods, Anderson (1979) showed that the gravity model is also derived from Cobb-Douglas form
because within the class of traded goods, preferences are identical, expenditure shares for any good are
identical across countries. As such, the demand for 'i s tradable good in country j is ij i j jM Y
where i is the expenditure on country 'i s tradable good divided by total expenditure in country j on
tradables, j is the share of expenditure on all traded goods in total expenditure of country j and
( , ).j j jF Y N Because the value of imports of i plus domestic spending on domestic tradables is
equal to the value of exports of i plus domestic spending on domestic tradables ( ( ) ),i i j j i
j
Y Y
demand for 'i s tradable goods in country j can be expressed as /ij i i j j j j
j
M Y Y Y = i i j jY Y /
ij
i j
M . Although Anderson has successfully explained that the gravity model is based on Cobb-
Douglas utility expenditure function, his assumptions of the similarity of traded-goods preference, of
trade tax structures and of transport costs structures reveal limitation since these cases can not practically
applied. Therefore, there would be a tradeoff between bias and efficiency because countries have
considerable differences of income, preferences and production capacity, which lead to biased estimate
of GDP.
Helpman and Krugman (1985) nonetheless derived the gravity model from the increasing return to scale
models and showed that specialization and trade would persist even when countries had identical
relative factor endowment for a wide variety of models. They did not support the viewpoint that
comparative advantage explain the relationship in a gravity model because countries with similar level
of income have been shown to trade more. Helpman and Krugman saw this as evidence that these
countries are trading in differentiated goods because of their similarities.
Bergstrand (1985) has described the gravity model under the general equilibrium framework whereby
world trade activities are basically based on the constant-elasticity-of-substitution (CES) utility function
and profit maximizing agent behavior in countries assuming a single factor of production. For the
demand side, the CES utility function is expressed as:
1/
1/
1
jj
j
jt
N
j kj jj
kk j
U X X
(1)
1,...,j N
18
This term denotes constant elasticity of substitution. According to Anderson (1979), the CES traded goods utility indicator is 1/
j ik ijk
i k
U M
where ijkM is the quantity of good k from country i consumed in country .j Good k is a
different commodity in each country due to differentiation. The elasticity of substitution is 1/ (1 ).
59 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Where ( )kj jjX X is the amount of 'k s aggregate good ( 'j s domestically produced good) demanded by
'j s consumers, ( 1) /j j j where j is the CES between domestic and importable goods in j
and ( 1) /j j j where j is the CES among importables in .j And expenditures in j are
constrained by income:
1
N
j kj kj
k
Y P X
and /kj kj kj kj kjP P T C E (2)
Where kjP is the k-currency price of 'k s product sold in the 'j s market,
jkT is one plus 'j s tariff rate
on 'k s product, kjC is the transport cost factor to ship 'k s product to j and kjE is the spot value of 'j s
currency in terms of 'k s currency. Maximizing (1) subject to (2) generates ( 1)N N first order
conditions that are solvable for ( 1)N N bilateral aggregate import demand equations:
1
11/(1 ) 1/(1 )
1 1 1'' ''j j j
j jj j j
kj jjkjP P P
(3)
And N domestic demand equation:
1
11/(1 )
1 1''j
jjj jD
jj j jj kj jjX Y P P P
For the supply side, Bergstrand employed the profit function maximized by firms in each country i in
each year:
1
N
i ik i i
kk i
X W R
(4)
Where iR is the amount available of the single, internationally immobile resource in a given year in i to
produce the various outputs and iW is the i -currency value of a unit of .iR R in each country is
allocated according to the constant-elasticity-of-transformation (CET) joint production surface:
1/1/
1
ii
i
i i
N
i ik ii
kk i
R X X
(5)
Where (1 ) /i i i where i is 'i s CET between production for home and foreign markets and
(1 ) /i i i where i is 'i s CET for production among export markets. This specification allows
the elasticity of transformation of supply between home and foreign markets and that among foreign
60 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
markets to differ. Substituting (5) into (4) and maximizing the resulting equation yields 2N first order
condition that are solvable for ( 1)N N bilateral aggregate export supply equations:
1
( ) 11/(1 ) 1/(1 )
1 1 1' 'i i i
i ii i i iS
ij i ij ik ik iiX Y P P P P
(6)
And N domestic supply equation:
1
11/(1 )
1 1'i
ii i iS
ii i ii ik iiX Y P P P
Where with one factor of production, national income in i is constrained by i i iY W R
For equilibrium, assuming 2N equilibrium conditions: D S
ij ij ijX X X (7)
Where ijX is the actual trade flow volume from i to .j However, the reduced form of
ijX in this
system would only be a function of ,iR ijT and .ijC No endogenous exporter and importer incomes are
included as one important factor of the gravity model. Bergstrand (1987) then solved for the gravity
equation with a partial equilibrium approach under the three assumptions.
The first assumption is that the market for the aggregate trade flow from i to j is small relative to the
other 2 1N markets. The small market assumption implies that variations in ijX and
ijP to
equilibrate D
ijX and .S
ijX Combining one each of (3) and (6) with one of (7) yields:
1/1
/ 1 1 / 1/ 1 1 / 11 1' '' ' '' 11 1 11
i j
j j j j ji i i i ij jj ji i i i i
kj kjij i j ij ij ij ik ik ii jjP Y Y C T E P P P P P P
(8)
and
1/
/ 1 1 / 1/ 1 1 / 11 1' '' ' '' 11 1 1
i jij
i j j j j jj i i i i ij jj i j i j i j ji i i i
kj kjij i j ij ij ij ik ik ii jjX Y Y C T E P P P P P P
(9)
The second assumption includes identical utility and production functions across countries ensures that
parameters in (8) and (9) are constant across all country-pairing. Combining (8), (9) yields:
( 1)( )/(1 )( )( 1) / 1 1' ''1 / ( 1)/ 1 / 1 / 1 1
( 1)/( )( 1)/( ) 1 / 11 / 1 1' ''1 1 1
kji j ij ij ij ik
ij
kjik ii jj
Y Y C T E P P
PX
P P P P
(10)
61 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Where ijPX is the value of the trade flow from i to j and equation (10) is regarded as a generalized
gravity equation.
The third assumption indicates perfect substitutability of goods internationally in production and
consumption, perfect commodity arbitrage, zero tariff and zero transport cost which implies that
1ij ijC T and ijP P for all , 1,..., Ni j . Because , equation (10) symplifies to:
1/2 1/21/ 2ij i jPX Y Y .
Bergstrand (1989) derived the generalized gravity equation, using per capita income as a proxy for the
exporter’s K/L ratio, which depends on whether the gravity equation is estimated for a capital or labor -
intensive industry. He incorporated factor-endowment variables in the spirit of Heckscher-Ohlin (H-O)
and taste variables in spirit of Linder. By analyzing consumers’ behaviors which are based on their
maximization of utility and firms’ behaviors to maximize their profits on two factor endowments such
as labor and capital, that Bergtrand solved for reduced forms, summed up across all firms in industry A
in country i yields the “generalized “ gravity equation:
( 1)/( )
( 1)( ) ( 1)/( ) 1
( 1)/( )( 1)/( )
* * 1 1
( 1)(1 )/( ) ( 1)/( ) ( 1)/(
( ) ( )
( / ) ( ) (1 )
A A AA A A
A A A A A AA A A
A A A A A A AA A A A
K
Aij i KA LB KB LA
LB KB i i j j
Aij Aij ij
PX Y
K L Y
C T E
)
( 1)/( ) ( 1)/( )1/(1 1
1( / ) ( ) /
A
A A A A A A AA A
A
Ain Ain Anj Anj nj
n n
P C P T E
Where AijP is the value of trade flow from i to j in industry A and K
iY is 'i s national output in terms of
units of capital *( ).A BK
i i KA Ahi KB Bhih hY K B X B X
In a typical gravity model estimated, GDP of exporting and importing countries is a proxy of national
outputs expressed in terms of units of capital. Per capita GDP of exporting and importing countries is a
proxy of capital-labor endowment ratio. Geographical distance is a proxy of transport cost and ijA
includes tariff costs which are imposed on exports, bilateral exchange rate.
More recently, Deardorff (1995) has proved the linkage between the gravity equation and the H-O
model by the assumption that each country could in theory specialize in producing one good and if
importers do not have the demand for such good, then trade between the two countries would be zero.
Since the gravity equation implies the expenditure system attached to the theory that higher income
consumers are ready to pay more for capital-intensive products, then exporters would supply the
relevant items. Deardorff also found the gravity equation relevant to the H-O theory. This explains why
capital-intensive countries export less than average to low-income labor-abundant countries. Deardorff’s
explanation may seem implausible since countries do not always require a single good, but as their
preferences change, their requirements may also change.
62 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Almost all theoretical papers on gravity for bilateral trade treat the trade barrier of geographical distance
between two countries pair as a proxy of transport cost. A country contiguous to its destination market is
not always at an advantage for trade. In reality, transport costs are a function of fuel costs, types of
products, volume of trade and therefore variant. It is impossible to identify them in the fixed effect
model that controls for unobserved specific effects of country pairs that do not vary over time.
Supposing that, two countries with a common border were separated from the rest of the world and had
the same level of economic development then they would be expected to trade more. In most cases,
such a country does not only trade with only one, other interacts with the rest of the world. This explains
why two contiguous countries sharing the same pace of economic development trade less. Anderson
and van Wincoop (2003) suggested a multilateral resistance term through the replacement of
geographical distance and border sharing in the U.S-Canada case (McCallum, 1995). Such a multilateral
resistance factor implies that if two countries were isolated from the rest of the world on one island in
the ocean, far away from the next continent, bilateral average trade costs measured by the iceberg factor
might be low and this should guarantee a high trade volume between both countries.
However, in fact, the two “isolated countries” are surrounded by many large countries therefore
multilateral resistance is much lower and thus trade between the two countries is lower even if GDPs
and trade costs are assumed to be at the same level. Rudolph (2010) has employed the total production
index of the exporting country and importing countries belonging to the rest of the world to compute the
multilateral resistance of trading countries for 23 OECD countries. However, data on a country’s total
production index are not always available.
4.3. Data and methodology
4.3.1. Data description
Export value: The chapter uses data on Vietnam’s export trade19
with over 150 countries across five
continents over a period of 13 years between 1997 and 2009. Such data are obtained via WITS on the
website of the World Bank, which closely collaborated and consulted with UNSD. The UN Comtrade
database contains information on exports and imports broken down by commodity and partner country.
The data base comprises various relevant items such as gross export value20
calculated in thousands of
USD annually, partner countries by country name and country code from 1997 to 2009. Overall,
Vietnam’s total and continental export value gradually increased until the year 2008 and fell back
considerably in 2009 as a consequence of the global crisis. Specifically, Vietnam exports more to Asian
countries and some other developed markets whereas the African market has been comparatively
neglected. The country’s export value peaked in 2008 but declined to about 55 billion USD in 2009. It is
noteworthy that the top 15 importing countries account for an average of 80% of Vietnam’s total exports
19
Vietnam’s export values to its importing countries for 13 years were adjusted based on the country’s GDP deflator index and
the purchasing power parity conversion factor of Vietnam’s currency required to buy the same amount of goods and services
in the domestic market as a U.S dollar would buy in the Unites States.
20
Gross export value or net export represents the value of a country’s total exports minus the value of its total imports and is
used to calculate a country’s aggregate expenditure, or GDP, in an open economy.
63 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
in this period. Although the rank of the major importing partners fluctuated yearly, it includes the United
States of America, Japan, China, Australia, Singapore, Germany, France, and the United Kingdom.
Importantly, the top 15 highest income countries are also the largest importing countries as mentioned
above. In fact, only 25% on average of the highest income countries have the highest labor
productivity21
and are not Vietnam's largest markets.
Importing countries: Data on GDP (constant 2005 U.S dollars) adjusted to PPP are available on the
World Bank’s website. In addition, we have obtained data on the labor force of importing countries over
this period via the KILM - data available from the ILO. Simply, by dividing real GDP over labor force,
we were able to generate the importer’s productivity index. In regard to expenditure equilibrium,
countries with a higher level of income will be prepared to pay more for goods and services. Although,
GDP per capita might seem to be a more plausible proxy for a country's demand capacity than simple
GDP since it shows the relative performance of the country would seem to be a good indicator for
comparing one country with another in terms of productivity, this variable is not included in the model
simultaneously with covariate the labor productivity due to a problem of multicollinearity. Depending
on the level of GDP per capita of importing countries, imports would tend to be labor or capital
intensive goods. In addition to this, we have taken the annual database on the real exchange of over 178
countries from Bruegel Exchange rate data of Darvas (2012).
Data on geographical distance between Vietnam and its country partners is available via CEPII. The
data base consists of a three-letter country code, the distance measured for a given country pair’s capital
and other dummy variables showing if exporters and importers share a common border (contig), a
common colony (comcol) and whether countries are entirely surrounded by land (landlocked).
4.3.2. Model and methodology
This chapter examines the export trade flows between Vietnam and all importing countries across
continents not only by employing the economic difference indicator but also the conventional indicators,
namely the income of the importing country and the geographical distance between the country pair.
Responding to the theoretical framework of the gravity model, in this chapter, we supplemented the
model so that it now includes the following determinants:
10 2 3 4 1 2
3 4 5
ln ln ln( / ) ln ln lnijt jt jt ijt ijt i j k
j ij ij t t ijt
EX GDP GDP LB DGDP Exchangrate Distance ASEAN
Landlocked Border Colony T e
Where:
ijtEX denotes the aggregated exports in USD by Vietnam to each destination market j at time t ;
0 is the common intercept;
jtGDP represents the Gross Domestic Product of importing country j at time t ;
21
Author has calculated countries’ labor productivity index by dividing GDP adjusted at power purchasing parity and constant
at 2005 by country’s labor force.
64 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
( / ) jtGDP LB denotes the Gross Domestic Product Per Labor of importers j at time t ;
ijtDGDP is the difference between GDP per capita of Vietnam and that of its importing partner and is
calculated by ijtDGDP = it jt
it jt
GDP GDP
POP POP . This indicator measures the difference in terms of the
relative factor endowments between two countries and is zero where there is equality in the relative
factor endowments.
ijtExchangrate is the real exchange rate of Vietnam and importing country j at time t .
ijDistance represents the geographical distance between the capitals of the two countries and is a proxy
of the transport costs;
The value one is attributed to ASEAN as dummy variable where the importing country is an ASEAN
member or zero otherwise. This indicates that such export trade is ASEAN intra-export trade; the value
one is also attributed to a dummy Landlocked, where the importing country is surrounded by land or
zero otherwise; Colony is a dummy variable that equals one if the importing country ever colonized
Vietnam and zero otherwise;
T is a set of dummy variables to capture year-specific “fixed” effects; and ijte is a normally distributed
error term.
The chapter aims to investigate Vietnamese exports to the global market over a period of 13 years from
1997 to 2009. Panel data are relevant to account for the unobservable effects, characteristics of each
partner country. Hence, the purpose of employing panel data is to model exports to partner countries as
a function of the income and productivity of importing countries. These covariates vary depending on
partner country and time. Moreover, there are many time-invariant variables, which may affect the
independent variables such as colonial history, financial situation, religious affiliation and political
regime. Not accounting for country heterogeneity causes serious misspecification. Baltagi (2005)
considered that panel data are more informative, provide more variability, reduce collinearity of the
variables, give a greater degree of freedom and thereby improve the efficiency of the model. With the
additional, more informative panel data, more reliable parameter estimates can be produced. Individual
characteristics of importing countries rather than conventional determinants by importing countries can
nevertheless influence exports. For comparison purposes, the above proposed model is run using the
pooled OLS as follows:
0 ijt i jtijtY X e Where i = 1, 2…, N and t = 1, 2., T
And satisfied with e│ X~ iid (0, σ2
NI )
N T N T N T N T N TQ E E I I I J J I J J
This is because there is a strong assumption for the application of the pooled OLS model that 0
should be common to all countries meaning that there is no difference between specific importing
countries in terms of unobservable effects. If this condition is satisfied, pooled OLS is unbiased and
65 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
consistent. In other words, if there are some individual effects hidden within the residual 0ijt ijte e
or
α0 is correlated with ijtX then ' 1 '[ ] [ ) ]ijt ijt ijt iE b b E X X X
and pooled OLS becomes biased.
Once the regression shows unobservable effects, the pooled OLS gives biased estimates and then, the
fixed effect model is preferable. In the fixed effect model, α0 is thought of as a fixed parameter, which
enables dummy variables to be introduced for each exporter-importer pairs. These dummy variables
allow correction of this error term. Albeit, another problem would arise where the database contains
many country partners thus creating a need for many dummies. Estimated N+K parameters for the
model would cause a multicollinearity problem. According to Cheng and Wall (2005), the general
model of trade between country pairs could be characterized by
'
0ln kt t k kt kt ktX Z t = 1., T; (1)
Where ktZ is the 1× k row of vector of gravity model specified in detail in the equation (1) above. The
intercept normally has three parts, one which is common to all years and trading partners at the level of
country 0 , the other is specific to year t and common to all trading partners, t and one which is
specific to country pairs and common to all years. The disturbance k t is assumed to be normally
distributed with zero mean and constant variance for all observations. To avoid the multicollinearity
problem rooted in a large (N+T+K-1) matrix, the fixed effect estimates of can be obtained by
performing the following transformation proposed by Wallace and Hussain (1969):
N T N T N T N T N TQ E E I I I J J I J J
Where NE = NNI J and .T T TE I J This transformation removes the country and time specific
effect. In fact, y Qy has a typical element . .( ..)ijt ijt ij ty y y y y where .. /i t ijty y NT
and the regression of y Qy and X QX to obtain the estimator ' 1 '( )X QX X Qy .
One side effect of the fixed effect model is that the transformation simultaneously removes time
invariant variables or time invariant effect. It is not possible to identify the distance effect and the
dummies such as “ASEAN” or “landlocked”. Rather, the fixed effect estimator ignores variations across
the countries. The characteristics within individual countries may or may not correlate with the
explanatory variables. In such cases, the instrumental method is appropriate to treat the correlation
between some explanatory variables and unobservable effects.
In contrast with the fixed effect, the random effect could be an alternative solution if the variation across
entities is assumed to be random and uncorrelated with the predictor or independent variables included
in the model. If 2(0, ),
ii IID 2(0, )tt IID and
2(0, )ijt vv are independent of each other,
then this is a two-way random effects model. In addition, ijtX is independent of i and t and itv for all
i and .t From (1) one can compute the variance-covariance matrix
66 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
' ' ' ' ' 2( ) ( ) ( )i i i i t t t t v NTE uu E Z Z E Z I
= 2 2 2( ) ( ) ( )i N T t N T v N TI J J I I I
In order to obtain 1, NJ is replaced by ,NNJ NI by ,N NE J TJ by TT J and TI by T TE J and
collecting terms with the same matrices. This gives 4
1
i i
i
Q
. Each iQ is symmetric and idempotent
with its rank equal to its trace. Moreover, iQ are pairwise orthogonal and sum to the identity matrix.
The advantages of this spectral decomposition are that4
1
.r r
i i
i
Q
. Where r is an arbitrary scalar so
that 4
1/2 1/2
1
( / ) .v v i i
i
Q
And the typical element of * 1/2
vy y is given by
*
1 . 2 . 3 ..ijt ijt i ty y y y y
1/2 1/2 1/2
1 2 2 3 3 1 2 41 ( / ), 1 ( / ), ( / ) 1.v i v i v i As a result, GLS
can be obtained as OLS of *y on *,Z where * 1/2 .vZ Z
The implication behind random effect is that, all individual differences are captured by the intercept
parameters. In the random model, the intercept parameters consist of a fixed part that represents the
population average and the individual differences from the population average. In contrast to the fixed
effect model, by definition, the random effect approach allows the time-invariant covariates to be
estimated. A random effect estimator is a generalized least square procedure and the fixed effect is a
least square estimator. In large samples, the GLS estimator has a lower variance than the least square
estimator. However, a disadvantage of the random effect sample is that the assumption on the
relationship between unobservable effect and explanatory variables goes to zero and if any explanatory
variable correlates with unobservable effects, the estimates become biased and inconsistent.
4.4. Empirical results
The empirical results reported in Table 9 present the estimates using the pooled OLS, fixed effect and
random effect approach. The value of exports, GDP, GDP per labor force and the difference between
country’s GDP per capita in USD are shown in the form of natural logarithms. For efficiency, all
coefficients are corrected for standard errors. The pooled OLS model shows heteroskedasticity and
imperfect auto-correlation. In addition, a groupwise heteroskedasticity problem arises in the fixed effect
model implying that the variance within country pair exports is not homogenous. The Breusch and
Pagan Lagrangian multiplier test for random effects reveals heterogeneous variance between country
pair exports.
Pooled OLS will be biased and inconsistent if unobservable effects are found to correlate with
independent variables. Unobservable effects may be the other time-invariant factors that are not
included in the model such as political or social characteristics within each partner country. We
therefore implemented fixed and random effects regression controlling for all time-invariant differences
between importing countries then tested to choose fixed or random effects model of Hausman (1978).
Although the Hausman test does not reject the null hypothesis that the random effect test is appropriate,
67 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Table 9. Gravity model for export flows of Vietnam
(Dependent variable: export value in USD, robust s.e)
Independent variables Pooled OLS
(1)
Fixed effect
(2)
GLS random
(3)
LnGDPjt 0.951*** 1.925*** 0.976***
(27.33) (2.95) (14.65)
LnGDPjt/LBjt -0.027 -0.947 -0.0331
(0.58) (-1.43) (-0.29)
LnDGDPijt -0.300 -0.530 -0.461*
(-1.40) (-1.35) (-1.65)
LnExchangrateijt 0.955***
(5.39)
0.146
(0.45)
0.270
(0.87)
LnDistance -0.311*** . -0.264
(-4.10) . (-1.19)
ASEAN 1.876*** . 1.921***
(8.52) . (3.00)
Landlockedj -0.974*** . -1.039***
(-8.54) . (-3.31)
Borderij 1.917*** . 1.950***
(7.94) . (2.78)
Colonyi 0.747*** . 0.742***
(3.88) . (3.49)
Constant 2.083* -14.86 1.929
(1.67) (-1.24) (0.72)
Year dummies (12 years)
Observations
yes
1871
yes
1871
yes
1871
No of groups
R2
-
0.73
165
0.56
165
0.73
LM (p-value)22
Breusch-Pagan/
Cook-Weigsberg23
: 2 (1)
Wald test (p-value)24
Hausman test: 2 (15)
34.74***
85.13***
-
-
-
-
30435.4***
45.06***
-
-
-
-
Robust t-statistics in absolute value in parentheses, * significant at 10%; ** significant at 5%; *** significant at
1%.
22
Wooldridge test for autocorrelation in panel data with H0: no first-order autocorrelation
23
Breusch-Pagan / Cook-Weisberg test for heteroskedasticity with Ho: Constant variance
24
Modified Wald test for groupwise heteroskedasticity in fixed effect regression model with H0: sigma(i)^2 = sigma^2 for
all i
68 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
the results are not identical with Hausman’s two options which are sigmamore and sigmaless. The
former option is based on both (co) variance matrices on disturbance variance estimate derived from the
efficient estimator, the latter based on both (co) variance matrices on the disturbance variance estimate
derived from the consistent estimator. Hence, the Hausman test with the two options favors the fixed
effect model.
As expected, the coefficients for the income of importing countries are positive and significant at 1%
implying that higher income countries would import more from Vietnam. Apparently, the OLS method
underestimates the GDP coefficient because it does not take in to account of a possible relationship
between independent variables with unobservable effects. When holding other repressors remains
constant and income of Vietnamese trading partners therefore increases by one percent, export volume
by Vietnam to its importing countries would increase by 1.92%.
If the gravity model is type of a reduced expenditure equation, it explains why it confirms the theoretical
notion that higher income induces higher expenditure. Regarding to the utility function suggested by
Bergstrand (1985) where the total good and service utility of one country is subject to its income it
means that if that country is able to satisfy its demand for goods then it would expand its ability to
increase consumption of domestic and imported goods. If consumers have to make a trade off either
between these two categories between two certain kinds of goods or their quantity, the country utility
would be improved if incomes rises and the demand curve would shift to the right. Empirical studies on
country level data such as Zarzoz and Lehmann (2003), Papazoglou (2007), Porojan (2001), Carrère
(2006), Tu and Chu (2014), Zhang and Kristensen (1995), Frankel and Rose (2000), Filipi and Moloni
(2003) have also proved the GDPs contribution to the intensity of trade among developed and
developing countries.
Furthermore, gravity studies using firm-level and product-level data also highlight the important and
positive role of GDP such as Bastos and Silva (2010) and Nguyen (2009). On the finding of Nguyen
(2009), the income of importing countries determines export volume of particular product items. The
sensitivity of the coefficient GDP for heterogeneous products is higher than that of homogenous ones
implying that richer countries would absorb more products, which are heterogeneous.
GDP per labor force effect is negative across the country pairs export trade, indicating that an increase in
labor productivity of an importing country will cause the Vietnam’s exports to decrease. However, the
coefficients are not significant across the pooled OLS, fixed effects and random effects regressions.
Sandberg et al. (2002) discussed that GDP per capita by importing country would much better capture
absorptive capacity and sophistication of demand for imported goods since it is a measure of the capital
labor ratio in the importing country. However, not all citizens can enter the country’s labor market;
hence, income per potential worker is more accurate indicator to compare two countries with the same
level of income and population, which may have different productivity. In fact, if this term could
account for unemployed workers, the result would be more plausible.
69 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
In our case, it is not possible to obtain a good database for the unemployment rate due to a missing time
series of the relevant data. Under the quality sorting, Bastos and Silva (2010) find that income per
worker of a country may affect the quality of its import products and the competitiveness of firms in
exporting countries. As such, an increase in the importing countries’ labor productivity induces a growth
in export quantity and unit value on the part of exporting countries. As regards the high competition in
the importing countries, more qualified firms who serve export products could be very competitive in
the rival markets. This finding clearly shows the importance of export pricing in dealing with the
difference in purchasing power or low cost labor in the local production sectors. For country-level data
of exports, it is not possible to evaluate country’s labor productivity on the unit price. In the case of
Vietnam, if country partners are more productive, they tend to import less.
The coefficient, the difference in income per capita between Vietnam and its trading partners indicator is
found to be negative, indicating that the greater the difference in income per capita between Vietnam
and its trading partners, the lower its export volume. Meanwhile, the effect of the difference in income
per capita between exporting and importing countries is found to be somewhat ambiguous in the study
of Serlanga and Shin (2004) however, it yields a positive effect in the paper of Rault et al. (2007). The
positive finding confirms the Heckscher and Ohlin (H-O) theorem that countries will export products
that make use of their abundant and cheap factors of production. However, we find this indicator
negative and consistent with Marie (2009) and Nguyen and Heo (2009). According to Nguyen and Heo,
the negative effect of the difference in GDP per capita between Vietnam and its 23 trading partners25
fits
the Linder hypothesis. As noted, Linder (1961) conjectured that the international trade pattern is
determined by the demand structure of countries. That is to say, the more similar the demand structure
of countries, the more they will trade with one another. Furthermore, international trade will still occur
between two countries having identical preferences and factor endowments. The findings of Baltagi et
al. (2003), Egger and Pfaffermayr (2003) advocate new ideas on the international trade theory including
Linder’s hypothesis of preferred model with full interaction effect26
and the classical H-O model in
regard to incomplete interaction.
Basic factor inherent to the gravity model is geographical distance between two trading countries, which
is a proxy for transport costs or trade barrier, depressing bilateral trade. Empirical testing for the effect of
distance is very popular in a wide range of gravity papers such as Longo and Sekkat (2004), Sohn
(2005) and Elliot (2007), Tu and Chu (2014), Silva and Tenreyoro (2006). Effect of distance within
export trade by Vietnam and its over 150 countries is somewhat ambiguous among the regressions. For
the OLS and model, it is negative and significant while in random effects regression, it is still negative
but insignificant. The problem is that distance does not change over time while transport costs such as
25
Those trading partners comprises 6 ASEAN country partners and 17 non-ASEAN countries over a 16-year period from
1990 to 2005.
26
Full interaction effect manipulated by Baltagi, Egger and Pfaffermayr includes exporter- by- importer, exporter- by- time
and importer- by- time interaction effects. In this study, coefficients of GDP per capita difference by country pairs give positive
sign in incomplete interaction effect. For full interaction effect model, this sign is counterbalanced. The paper has tested a
generalized gravity model by bilateral export volumes of triad economies namely EU15, USA, Japan and their 57 most
important trading partners between 1986 and 1997.
70 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
fuel cost, or any fee applied to certain export goods are variant over time. Consequently, Brun et al.
(2005) have treated distance as a function of oil price and infrastructure variables. Thereby, they
demonstrated that the effect of distance does not diminish over time even though this effect changes
from negative to positive27
.
There are several reasons why geographical distance is not always a good proxy for trade obstacles.
Firstly, this covariate does not change over time and it can not be identified in the fixed effect model that
captures unobservable effects within each country pair. Secondly, the result in dubious estimates where
two states are distant from one another but their mutual trade more if their economic sizes are explicitly
large. Thirdly, the level of mutual trade that two contiguous countries would trade more or less depends
on their interconnection with the rest of the world. Fourthly, foreign trade rests on how much domestic
trade within exporting or importing countries take place. If domestic trade accounts for most of a
country’s production capacity, its export volumes would fall correspondingly. It is interesting to confirm
that distance remains a hindrance to trade in empirical findings for a particular case of Vietnam
(Nguyen, 2010). Critically, the authors have treated their model as fixed and random for the former and
dynamic GMM.
Anderson and van Wincoop (2003) have proposed a MRT28
as another term for trade hindrance. Trade
cost in the context of the MRT considers country’s trade with the rest of the world, which includes the
cost of intra-national and international trade and the CPI. The general equilibrium structure with many
countries trading a variety of goods differentiated by country origin is expressed as follows:
1.
.( ).
i j ij
ij
w i j
Y Y tX
Y P P
27
If the regression is manipulated by OLS based on separate years (1997-2009), effect of distance is negative and significant.
The coefficients are not fixed and fluctuates in a range [-0.75: -0.48].
28 Starting from the idea of utility function, Anderson and Van Wincoop assumed that if ijc is consumption by region j
consumers of goods from region i , consumers in region j maximize (1 )/ ( 1)/ /( 1))i ij
i
c subject to the
budget constraint ij ij j
i
p c y Where is the elasticity of substitution between all goods, i is a positive distribution
parameter, jy is a nominal income. If ip denotes the exporter’s supply price and ijt is the trade cost factor between i and j ,
then ijp = .i ijp t For each goods shipped from i to j , and if the exporter incurs export cost equal to 1ijt of country i
goods , the sum of value of production of the origin i ijp c and the trade cost ( 1)ij i ijt p c that the exporter passes on to the
importer. And total income of country i is therefore .i ij
j
y x As such, the nominal demand for country i goods by
country j , consumers satisfy by
1
i i ij
ij j
j
p tX y
P
Where jP is the consumer price index of j given by
1/1
1
.j i i ij
i
P p t
71 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Where iY and jY are total income of country i and country .j wY is the GDP of the whole world.
ijt is
the trade cost incurred by the exporter and or importer and is assumed to be symmetric ( ).ij jit t If
trade cost inside the exporting country is assumed to be 1 ( 1),iit then the trade cost that country i
passes on to country j is 1ijt and transport costs between the designate two countries are .ij ji it t P
where jP denotes the exogenous multilateral resistance of the exporting and/or the importing country.
For the MRT, if the multilateral resistance iP is high, this means high trade barriers for country ,i
demand for country 'i s goods is lower. In contrast, higher trade barriers for country j characterized by
high resistance ,jP higher demand for country i ’s goods. The positive sign of jP is in line with the
findings of Rahman (2003) and Rudolph (2010) fitting the theoretical view that the rise in importing
countries’ trade cost shift demand toward the international market. Because of the unavailability of the
intra-national trade data, we could not have applied the method postulated by Anderson and van
Wincoop (2003). Instead, we have exploited the real exchange rate as a proxy for the time-varying
component multilateral resistance factor of importing countries as proposed by Adam and Cobham
(2007). Because it is not able to observe directly iP and jP , the CPI or bilateral exchange rate of
countries could be a good reference. Although the regression result in Table 9 shows positive signs of
the exchange rate for OLS, fixed and random effect model, the significant effect is only found for the
OLS regression. The result implies that the change in exchange rates of Vietnam currency
insignificantly supports for the country’s exports in the period. To express in other word, we do not have
sufficient evidence to conclude that the depreciation of Vietnamese currency helps increase the
country’s exports.
The regression result shows that in the intra trade context of ASEAN, export value is significantly
positive, confirming the important role of the ASEAN in stimulating Vietnamese exports. The ASEAN
coefficient indicates that Vietnam’s membership of the ASEAN group has contributed to a 582%
increase in export volume between 1997 and 2009. However, the effect of the ASEAN dummy is not in
line with Nguyen (2010), Nguyen and Heo (2009) where the negative sign of the ASEAN dummy raise
a suspicion of its erosion. This discrepancy could be explained by the type of model used. While
Nguyen (2010) examined the ASEAN effect for trade between Vietnam and the other ASEAN
countries, Nguyen and Heo (2009) observed an ASEAN negative sign, which he explains as being to a
larger share of Vietnam's exports going to non-ASEAN countries. In fact, Vietnam has become deeply
integrated into the global market and thus less involved into intra-regional trade. The negative sign result
may also be partly due to the random sample taken for 15 largest trading partners of Vietnam. For our
export model, the ASEAN effect remains important. Although geographical distance is not important
for neighboring ASEAN markets, however, what is important for Vietnam is the effect of trade
liberalization through tariff and non-tariff removal between ASEAN country trading partners implying
that Vietnamese export flows to ASEAN countries largely rest on the tax relief regime.
Beside the geographical distance indicator, another trade barrier term is landlocked dummy. The
negative sign of the landlocked dummy indicates that importing landlocked countries find it a greater
72 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
challenge to absorb Vietnamese export products. The reason may be that over 90% of export contracts
are FOB29
, which are less favorable for landlocked countries. The result in table 9 suggests that median
value of Vietnamese exports to landlocked countries is about 182%30
lower than non-landlocked
countries.
Countries who share a common border with Vietnam still confirms their supportive factor for trade of
Vietnam. Especially, among these contiguous countries, China is emerging as huge producer and
advantageous of economies of scale since this is a very populous country which has a high demand of
goods consumption. A very significant positive effect of the variable “Border” shows that export
activity between Vietnam to these countries is very stimulated. Moreover, country who ever colonized
Vietnam namely France remains a very active importing country for Vietnam’s exports in the period.
4.5. Conclusion
Empirical studies using the gravity model have successfully explained the international trade pattern of a
country in that trade between a country pair is shown as proportional to their income level and inversely
proportional to the distance between them.
This chapter specifically considers export trade with all countries across the continents in the purpose of
examining Vietnamese export activity in the period 1997 to 2009. Among three types of regressions,
fixed effects model is found to be the most relevant. Significantly, the 15 largest importing countries of
Vietnamese goods in this period dominate the export market, accounting for some 80% of the annual
volume of exports. Nevertheless, high income countries remain important trading partners of Vietnam.
The augmented gravity model highlights the fact that high-labor productivity countries are not important
to Vietnam’s export trade and the difference in country’s per capita income strengthens the Linder
theory, suggesting that Vietnam should not continue to use take use of its low labor cost advantage.
Unlike in previous literature on Vietnamese trade, geographical distance is shown not to be a good
proxy for trade cost although its sign is negative. The positive ASEAN sign highlights the role of
regional economic integration where tax cuts help encourage imports from Vietnam by other ASEAN
members with a rejection probability of 1% significance. In particular, the positive role of contiguous
countries in stimulating export trade from Vietnam is still confirmed indicating that the border trade
effect is important.
The following chapter investigates and compares the role of individual firms in determining export trade
pattern in footwear, rice and wood and wood products sectors. Again, the combination of micro and
macro level data in a gravity model aims to find the effects of control variables, which influence
exporting firm's activity at the same level. However, since firms are dynamic and differentiated in terms
29
FOB stands for Free on Board and this term is applied to sea transport export contract
30
Technically, the figure is obtained by taking the antilog (to base e) of the estimated dummy coefficient then subtracting 1
(*100). For further detail, see Gujarati (2004, 321).
73 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
of resources and motivation to trade then effects of importing countries are expected to be different from
the country’s aggregate exports as well as among those export sectors.
74 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
CHAPTER 5. THE ROLE OF FIRMS IN DETERMINING EXPORT TRADE PATTERN:
VIETNAM'S FOOTWEAR, RICE AND WOOD AND WOOD PRODUCTS SECTORS
SUMMARY
This chapter investigates the role of firms active in the footwear, rice and wood and wood products
sectors to determine their trade pattern using the gravity model. The two main sources of data are
Vietnamese Customs Office for firm's export performance and the World Trade Indicator for gravity
factors. This chapter differs from the previous one in that it examines the effect of gravity factors on
Vietnamese firms' export performance for the three export sectors mentioned. Hausman test advocates
the fixed effect model therefore confirming the role of firms in determining their export trade.
75 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
5.1. Introduction
In recent times, the need to satisfy domestic consumption has led to an unprecedented rise in
international trade. Since every country has its own advantages in producing certain products,
international trade is a significant catalyst for local producers and foreign consumers. The classical and
new trade theories constitute the two main strands of the theory on international trade. While the
classical trade theory for absolute and comparative advantage has some limitations, such as the
economic models are simplistic in that only two trading countries and two products go into the model
whereas the new trade theory is superior taking into account many more countries and factors, which
help describe more explicitly the international trade pattern. A typical model of the new trade theory is
the so-called gravity model in which GDP and geographical distance as the two main determinants of
international trade.
In the gravity model, the GDP factor in most cases is a driving force for stimulating trade among
countries. By nature, it is a component of a pure expenditure system model where the consumption of
one good in one country is thus equivalent to the income of that country. Nonetheless, this expenditure
system model can only be applied in practice if it is assumed that income must equal sales. In fact, a
country does not spend all its income on imported goods from another country. Bergstrand (1985) has
proved by deriving the utility function that the gravity equation is a reduced form derived from a partial
equilibrium subsystem of a general equilibrium model with nationally differentiated products, since the
expenditure of a country is constrained by its income.
Empirical literature on the gravity model describes the role of the national income of a country as
encouraging bilateral trade giving a positive sign in the regression models. There is a long list of gravity
model papers on the subject of the country's total export and/or import trade which are treated as an
endogenous variable with GDP acting as an exogenous one. Moreover, products exported or imported
are expressed as total value and in turn reflects the aggregate value of trade and are positively affected
by the GDP of a country partner. Nonetheless, the role of firms in determining trade pattern has been
totally ignored in gravity model literature for the case of Vietnam.
Krugman (1980) a pioneer of the New Trade theory, identified that there is an incentive to concentrate
production of a good near its largest markets since this would enable the seller to minimize
transportation cost. His suggestion on the role of the transportation cost was a milestone of the gravity
model where the geographical distance between exporting and importing country typically is a negative
indicator. In reality, since firms are heterogeneous and some firms are more competitive than others, the
more productive firms may decide to take part in the international markets (Melitz, 2003). In some
recent empirical gravity papers such as Bernard and Jensen (2007), Arkolakis and Muendler (2009),
Bastos and Silva (2010), firms' heterogeneity and productivity have been shown to be crucial factors
determining their trade pattern. Since the total export and import trade of a country can be expressed as
firms' extensive and intensive margins, they enable us to examine how the gravity factors affect firms’
export performance.
76 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
In this chapter, we use data on firms' exports from Vietnam of footwear, rice and wood and wood
products sectors to over 150 countries across continents to investigate how firms determine their export
trade pattern. For comparison purpose, we revisit gravity model with similar indicators that have been
included in the previous chapter. However, five dummies of distance variable are included because we
detect that this variable turns out to affect positively those firms’ exports. Furthermore, the level of
mutual trade between two contiguous countries is also dependent on their extent of trade with the rest of
the world. In some cases, the scale of foreign trade rests on how much domestic trade takes place within
exporting or importing countries. If domestic trade accounts for most of a country’s production capacity,
its volume of exports would fall accordingly.
Although, empirical papers using country-level data have contributed to international trade, in some
cases, their results may be biased since the aggregate export trade data do not accurately reflect the role
of individual firms. Another factor is that exports by firms of in various sectors are heterogeneous. Some
firms serve a wide variety of markets while others serve only one. This may be because some markets
preferentially consume a certain kind of product, which satisfies the different taste of local consumers.
It is therefore clear that firms from the footwear, rice and wood and wood products sectors perform
differently in the international market. Among those sectors, rice products are considered homogeneous
therefore it is not appropriate to express total rice exports as the intensive margin of a product per firm.
We therefore regressed potential gravity factors on the export value and volume of firms which we then
applied to all three firms sectors. Notably, the national income is consistently and positively linked with
firms’ export value and volume across the three sectors. The signs of GDP per labor unit of importing
countries are dissimilar between these sectors, being positive for the footwear firms sector but negative
for the rice and wood and wood products sectors. Interestingly, the effect of the ASEAN dummy is not
consistent, changing from positive for the footwear and rice firms’ exports to negative for the wood and
wood products firms’ exports. The landlocked country dummy variable is more sensitive to the volume
of rice than its value, probably because rice is a homogenous and heavy product and its weight affects
the transport cost, which explains why landlocked countries are commonly an obstacle to imported
goods.
The structure of the chapter is organized as follows: Section 5.2 demonstrates the heterogeneity of
footwear, rice and wood and wood products firms active in international market, section 5.3 briefly
analyzes the dynamics of these firms, section 5.4 investigates their important importing countries, sector
5.5 presents the model and data, section 5.6 shows the role of firms' intensive and extensive margins in
their export performance, section 5.7 reports empirical results of the regression model and section 5.8
concludes the chapter.
5.2. Firm heterogeneity by market value
In order to clarify the role of firms in determining their export trade pattern, we investigate their activity
levels corresponding to the number of markets served in footwear, rice and wood and wood products
sectors. The term total firms i.e. "all firms active in the three sectors" in Table 10d is used solely for
reference and firm heterogeneity analysis is based on data for the individual sectors. Table 10a shows
77 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
that in general, footwear firms are the most dynamic for this period since a firm from this sector was
able to serve a maximum of 99 markets in 2008 while rice and wood and wood products firms are much
less dynamic with 49 and 48 markets served respectively in the same year. Moreover, the average
number of markets served by a footwear firm was the largest in 2008, reaching 9.59 markets while in
the same year a single rice firm averaged 5.3 markets.
Significantly, the main difference between firms in the three sectors was the number of market served
(ranging from one to over ten). A common feature of all three sectors is that the percentage of firms
serving a single market is high but the value of their exports was comparatively low. Conversely, firms
serving the largest number of markets accounted for the highest share of the value of exports. In the
footwear sector, there are 51.79% of firms serving only one export market and this percentage steeply
decreased corresponding to the increase in the number of market served.
However, since 2008, the percentage of firm serving a single market decreased reflected in the
following years in a higher percentage of firms serving more than one market. Firms performed more
dynamically in the following years as shown by the greater number of market served but that does not
necessarily mean they have become more productive or that the value of their exports increased.
Compared with the footwear sector, the percentage of firms in the rice sector serving a single market
was lower at 45.9%. In 2006, unlike the footwear sector, the percentage of firms serving three or four
markets remained constant. However, there is a rising trend in the percentage of firms serving only one
market. Although, those firms serving more than ten markets accounted for the highest value of exports,
the value of exports of those firms serving four, five and six markets were significantly higher than
firms of the footwear sector. In particular, in 2006 and 2007, the percentage of the value of rice exports
for firms serving six and four markets were significantly higher than for other firms, reaching 32.41%
and 45.8% respectively.
Surprisingly, wood and wood products firms performed very differently compared with their footwear
and rice counterparts. Although the percentage of firms serving a single market was the highest, the
percentage of the value of exports was distributed evenly across various categories of firms serving a
different number of markets but in particular, for those firms serving two, three or four markets, that
percentage was rather high compared with the footwear sector. In 2007, for the category of firm serving
a single market, the value of exports was extremely favorable, accounting for 23.25% of the total.
In short, the main difference between firms in these three sectors is that firms serving the international
market are very dynamic which may be one of the reasons why firms are able to influence the pattern of
Vietnam's export trade.
78 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Table 10. Firm heterogeneity by sector (2006-2010)
Table 10a. Footwear sector
Number of markets served 2006 2007 2008 2009 2010
% of firms % of value % of firms % of value % of firms % of value % of firms % of value % of firms % of value
1 51.79 0.46 48.26 0.50 41.75 0.33 45.53 0.47 41.34 0.32
2 9.42 1.03 8.19 1.73 9.32 0.17 8.75 0.22 8.01 0.35
3 4.71 0.19 6.10 0.20 6.80 0.21 6.81 1.11 5.43 0.25
4 3.41 0.51 3.31 0.22 3.69 0.20 2.72 0.16 2.58 0.06
5 1.62 0.54 3.14 0.60 2.91 0.47 3.31 0.74 2.33 0.48
6 1.79 0.45 1.74 0.58 2.72 1.48 2.14 0.42 2.07 0.79
7 1.62 1.09 1.05 0.17 1.94 0.28 2.33 1.07 1.55 0.27
8 1.46 0.39 1.57 0.45 2.14 0.96 1.36 0.44 2.58 0.51
9 1.3 0.39 1.92 0.87 0.58 0.44 0.78 0.12 0.52 0.01
10 1.62 2.17 0.87 0.23 0.58 0.90 0.97 0.43 1.29 0.98
More than 10 21.27 92.74 23.87 94.39 27.57 94.50 25.29 94.78 32.30 95.92
Average number of markets per firm
Maximum number of markets per firm
Number of firms
7.69
85
127
8.63
95
140
9.59
99
133
8.94
78
118
11
81
121
Table 10b. Rice sector
Number of markets served 2006 2007 2008 2009 2010
% of firms % of value % of firms % of value % of firms % of value % of firms % of value % of firms % of value
1 45.90 0.40 51.79 1.28 37.01 0.85 46.56 0.50 50.86 2.60
2 10.66 0.61 8.04 2.27 14.29 0.79 11.34 1.09 9.97 1.49
3 10.66 2 6.25 0.53 7.14 0.32 4.86 0.54 5.84 1.23
4 6.56 1.23 8.93 45.80 5.19 0.87 3.24 0.31 4.47 1.35
5 4.10 2.48 2.68 2.72 5.84 1.29 5.26 1.01 4.81 1.15
6 4.10 32.41 8.04 18.38 5.19 17.27 4.45 1.36 3.09 1.44
7 1.64 4.48 3.57 5.03 3.90 2.78 2.43 1.31 4.12 2.98
8 2.46 1.71 3.57 3.92 1.30 0.98 2.02 8.78 2.41 2.02
9 0.82 10.66 0.89 2.17 1.95 0.93 3.24 51.68 2.06 2.56
10 4.10 5.75 1.79 2.09 4.55 3.08 2.02 1.51 1.72 1.84
More than 10 9.02 38.21 4.46 15.77 13.64 70.78 14.57 31.84 10.65 81.28
Average number of markets per firm
Maximum number of markets per firm
Number of firms
4.24
41
86
3.56
30
72
5.33
49
106
4.90
43
125
4.15
41
115
79 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Table 10c. Wood and wood products sector
Number of markets served 2006 2007 2008 2009 2010
% of firms % of value % of firms % of value % of firms % of value % of firms % of value % of firms % of value
1 52.01 9.21 53.60 71.34 47.58 16.60 52.63 11.50 53.44 12.31
2 13.94 9.29 15.55 6.42 15.71 9.32 15.06 10.91 14.73 10.28
3 7.87 6.23 7.98 3.08 8.86 6.35 7.33 10.76 7.67 11.35
4 5.57 4.92 5.86 2.16 5.69 5.12 5.95 8.76 4.69 6.63
5 3.36 6.13 3.47 1.44 3.63 5.88 3 3.78 3.70 6.41
6 2.96 5.44 2.95 2.12 3.12 5.35 2.59 5.04 2.86 4.72
7 1.96 5.59 2.28 2.57 2.11 2.81 2.02 5.39 2.14 5.42
8 1.20 4.51 1.76 1.61 1.31 2.48 1.86 4.75 1.76 4.06
9 1.40 5.40 1.40 1.18 2.17 3 1.50 2.06 1.26 3.94
10 1.45 3.30 1.40 1.36 1.76 3.35 1.05 2 1.15 2.41
More than 10 8.27 39.93 3.73 6.66 8.06 39.68 7 35 6.60 32.42
Average number of markets per firm
Maximum number of markets per firm
Number of firms
3.43
39
127
2.88
38
114
3.64
48
113
3.24
43
137
3.17
37
135
Table 10d. Total
Number of markets served 2006 2007 2008 2009 2010
% of firms % of value % of firms % of value % of firms % of value % of firms % of value % of firms % of value
1 48.58 2.93 50.02 23.25 44.57 4.92 49.17 4.13 50.60 12.12
2 13.05 3.15 13.89 2.71 14.36 2.79 14.16 4.05 13.57 3.37
3 7.60 2.28 7.81 1.81 8.30 1.92 7.20 4.15 7.39 3.56
4 5.40 1.92 5.60 9.42 5.43 1.75 5.57 3.06 4.58 2.18
5 3.40 2.53 3.47 1.28 3.93 2.21 3.32 1.81 3.76 2.26
6 2.96 7.81 3.15 4.47 3.19 6.79 2.70 2.22 2.84 1.94
7 2.20 3.23 2.41 1.87 2.24 1.59 2.08 2.45 2.37 2.24
8 1.36 1.77 1.84 1.48 1.57 1.46 2.02 4.22 1.93 1.81
9 1.40 3.80 1.63 1.22 1.93 1.28 1.43 14.77 1.29 1.70
10 1.76 3.19 1.35 0.94 1.61 2.17 1.30 1.97 1.33 1.50
More than 10 12.28 67.32 8.83 51.51 12.86 73.06 11.04 57.11 10.35 67.28
Average number of markets per firm
Maximum number of markets per firm
Number of firms
4.74
85
340
4.38
95
326
5.06
100
352
4.43
78
380
4.29
81
371
Source: Author’s calculations based on VCO database
80 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
5.3. The dynamics of firms by export criteria
The important export factors presented and compared across sectors in Table 11 are the number of
export transactions, export firms, destination markets and product types from 2006 to 2010. Wood and
wood products sector emerged as the most dynamic and increased rapidly the number of export
transactions year on year, followed by the footwear and rice sector. Significantly in 2009, when the
Vietnamese economy in general was affected by the world financial crisis, the number of export
transactions of the rice and wood and wood products sectors strongly increased while decreasing in the
footwear sector. The number of firms active in international market varies between sectors reflecting
varying demand for different products from foreign consumers.
In fact, there are a larger number of wood and wood products firms compared with the other sectors
whereas there are far fewer rice export firms mainly because the demand from developed countries is
rather low. Moreover, the dynamics of a firm are shown by the number of markets it served assuming
that the destination markets for each sector are relatively equal.
We use the concept of product heterogeneity across the sectors to determine a firm's ability to export
different products since heterogeneous products supplied by firm in a sector in international market
indicates the degree of a firm's differentiation. The wood and wood products sector has the highest
level of product differentiation and rice the lowest. However, this large difference in product
differentiations between rice and wood and wood products sectors does not necessarily mean that the
former is more competitive than the latter since rice is typically homogeneous product.
81 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Table 11. Exporting firms’ criteria across sectors
Export
criteria
Year
Sector
2006
2007
2008
2009
2010
No o
f ex
port
tran
sact
ion
s
Footwear
335139 466833 615495 568455 823536
Rice
4073 3793 3577 7650 10117
Wood and wood products
441892 696211 602218 881423 1023183
No o
f fi
rms
exp
ort
ing
Footwear
671 682 980 682 511
Rice
144 134 170 270 324
Wood and wood products
2233 2360 2208 2734 2976
No o
f
des
tin
ati
on
mark
ets
Footwear
161 168 172 141 146
Rice
99 82 127 150 149
Wood and wood products
168 160 149 178 165
No o
f p
rod
uct
typ
es
Footwear
764 460 405 363 128
Rice
50 41 8 23 41
Wood and wood products
1085 915 494 611 598
Source: Author's calculations based on VCO database
82 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
5.4. Largest importing countries by export value
The Figure 18 shows from 2006 to 2010, United States of America (USA) and Britain were the two
largest import markets for footwear from Vietnam. Out of the top 10 importing countries, Mexico
imported the least. Export revenue from the USA and Netherlands markets strongly fluctuated over the
period whereas for other countries, it was relatively constant. The value of Vietnamese exports to USA
and Netherlands peaked in 2010 and 2008 respectively. In 2010, Vietnamese footwear revenue from the
USA amounted to 1,443 million USD, a two-fold increase compared with 2006. Vietnamese footwear
exports to the remaining countries fluctuated slightly though the period. Specifically, export revenue of
Vietnam's footwear sector to the ten largest importing countries decreased significantly in 2009
compared with 2008. It should be noted that for the USA, UK and Netherlands, in 2010, the export
value rose dramatically to almost 1440 million USD, 515 million USD and 442 million USD respectively
compared with 2009.
Of the top 10 rice importing countries, the Philippines was the largest importer accounting for over 50%
of the Vietnam's total rice export revenue in 2008 and remained constant until 2010. Malaysia was the
second largest rice importing country with the peak of 271.2 million USD in 2009, followed by Cuba,
which reached 400.1 million USD in 2008. Indonesia was the fourth largest importing market for
Vietnamese rice firms, however, its demand fluctuated strongly over the period.
The USA continues to confirm as the largest importing country for the wood and wood products sector
of Vietnam in the 2006-2010 period. Japan was the second largest importing country with 424.8 million
USD for importing these products in 2010. Being different from the footwear and rice sector, export
value of the USA increased strongly in 2009 meanwhile except for China and France, export value of
the remaining importing countries increased slightly in the same year. However, in 2007, except for
China, there was a sudden decrease in exports by Vietnam's wood and wood products firms to these
countries. Especially, in the same year, export value of Vietnam's wood and wood products firms to the
USA fell sharply by 32% compared with the previous year.
83 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Figure 18. 10 largest importing countries by export value (mil. USD)
Source: Author's diagram based on VCO database
0
200
400
600
800
1000
1200
1400
1600
Footwear
2006
2007
2008
2009
2010
0
200
400
600
800
1000
1200
Rice
2006
2007
2008
2009
2010
0
200
400
600
800
1000
1200
1400
Wood and wood products
2006
2007
2008
2009
2010
84 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
5.5. Methodology and data
5.5.1. Gravity model and firms’ trade
Another strand of empirical literature on the gravity model focuses on firms' extensive and intensive
margins just derived from the total exports or imports of a given country. This chapter however, differs
from existing literature in that it examines how firms' exports in a particular export sector are influenced
by the economic factors of the importing country market. In addition, firms are able to diversify their
activities within a particular market depending on changes in demand of importing countries. A
conventional gravity equation where the income of both exporting and importing countries are included
would capture the parallel effect of partner countries' economic size on bilateral trade. In the case of a
single exporting country and many importing countries, the exporting country's income is captured in
the constant term of the gravity model regression.
As one important factor of the gravity model, GDP is normally a proxy for the economic size of a
country. However, GDP per capita is a better measure since it represents the output per person of a
country. The theoretical framework previously described shows that demand for imported goods in any
country would be constrained by its budget and a higher national income would induce a larger demand
for goods. In this chapter, we regress export value and volume of firms with gravity factors which were
applied in the previous chapter namely GDP, GDP per labor force and the difference between GDP per
capita of Vietnam and its importing countries. Moreover, time-invariant variables are also included such
as ASEAN, geographical distance, landlocked, border and colony.
We examine the effects of the gravity factors on exports of Vietnam's footwear, rice and wood and
wood products firms through their extensive and intensive margins. This approach allows us to test the
role of heterogeneous firms engaging in country pair trade. Nevertheless, if we consider the total exports
of a country, we would be ignoring the dynamics of firm therefore probably causing the anomalous
results. Firm's intensive and extensive margins can be obtained by decomposing the total export and/or
import trade of a particular sector. We apply the margins proposed by Arkolakis and Muendler (2009)
so that total exports sdT from source country s to destination country d can be decomposed into
sdsd sdT M t where sdM is the number of exporters (intensive margin) in 's with shipments to
destination d and /sd sd sdt T M are these exporters mean sales to d (extensive margin).
Alternatively, in most cases where the total exports relate to all export varieties, total exports sdT can be
decomposed into: sdsdsd sdT M G z where sdM is the number of exporters in s with shipments to
destination ,d ( )sd
sd dG G
is the total number of products exported from s to d and
/sd sd sdz t G is the average value of exports per product per firm (Bernard et al. 2007a).
As stated earlier, since this chapter only tests the gravity model for three particular sectors, the extensive
margin is the average export sales by a firm to its destination markets. Remarkably, as shown in Figure
85 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
19, the total value of exports by footwear, rice, wood and wood products firms (extensive margin) has a
strong positive relationship to the number of firms and the number of markets (intensive margin).
Figure 19. The correlation of firm's extensive and intensive margins
Source: Author's diagrams using STATA 12.0
implying that both firm’s extensive and intensive margins play an important role in determining the total
value of exports.
The graph shows, heterogeneous firms who were able to diversify their markets and maximize their
exports sale were also more productive than others. However, since firms had to compete with the
rivals, only a few firms were able to dominate in the foreign markets.
02
46
02
46
-10 0 10 20 -10 0 10 20 -10 0 10 20
2006 2007 2008
2009 2010 Total
lnnu
mfir
m
lntotalexportGraphs by year
01
23
01
23
-10 0 10 20 -10 0 10 20 -10 0 10 20
2006 2007 2008
2009 2010 Total
lnnum
mark
et
lntotalexportGraphs by year
86 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
5.5.2. Data
Micro data: Export data of firms in the three sectors for a period from 2006 and 2010 were obtained
from the VCO. This government-based body is responsible for managing export and import activities of
firms in Vietnam as well as collecting data on their exports and imports. Firms who have the need to
export or import goods are required to make a declaration to a border gate customs sub-department of
the VCO. Such declaration includes date of application, firm identity code, importing country code,
transaction code, currency code, exchange rate, export volume, unit price, and export value in the
foreign currency31
, name of product and its code at 10-digit SITC level.
We only used firms' data for export transaction in USD although in fact, 15 different currencies were
used by firms. On average, the value of export transaction in USD for footwear, rice, wood and wood
products correspond to 99.2%, 92% and 95.2% of the total value. For the export volume dataset, we
only kept the following units "pair"32
, "ton"33
and "piece"34
for regression.
Macro data: We obtained data on the following gravity factors: constant GDP of the importing
countries (measured at PPP) are available from WDI of the World Bank's website. In addition, we have
obtained data on the labor force of importing countries over this period via the KILM - data available
from the ILO. Real exchange rate data as presented in chapter 4 are taken from Bruegel website. The
data on geographical distance between Vietnam and its importing countries and the remaining variables
were extracted from CEPII.
5.6. Gravity factors and export margins
Regression results of three firms’ export margins (total exports, total weight and total number of firms)
for three export sectors are presented in Table 12, 13, and 14. The signs of GDP are consistently positive
compared with those of gravity regression model result in chapter 4 using country level data.
Surprisingly, the signs of GDP per labor force are very different across three firm sectors. This indicator
is significantly positive for the footwear firms sector whereas it shows a negative sign in all margins of
the rice firms sector and in the wood and wood products firm sector, GDP per labor force gives a
positive sign which is only significant for the number of firms margin. As such, high-income countries
present as very attractive destination markets for footwear firms but the effect of GDP per labor force on
the rice sector margins runs counter to this stylized fact that high income countries tend to import less
rice.
In contrast with the finding of the difference in GDP per capita between Vietnam and its importing
31
Payment currency in export contract includes AUD, CAD, CHF, CNY, EUR, GBP, HKD, JPY, MYR, NOK, SGD, TWD,
USD, USR, and VND.
32
This quantity unit accounts for 98.8% of the total
33
This weight unit accounts for 92.8% of the total
34
This quantity unit accounts for 96.8% of the total
87 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Table 12. Gravity factors and export margins (footwear firms)
(1) (2) (3) (4) (5) (6)
Independent
variables
Total exports
(ln)
Total exports
(ln)
Total quantity
(ln)
Total quantity
(ln)
Total number of firms
(ln)
Total number of firms
(ln)
LnGDPjt 1.028*** 1.036*** 1.001*** 1.046*** 0.965*** 0.977***
(18.96) (17.63) (15.72) (15.63) (18.74) (17.13)
LnGDPjt/LBjt 0.463*** 0.413** 0.518*** 0.437** 0.408*** 0.382***
(2.73) (2.32) (2.77) (2.20) (3.18) (2.85)
LnDGDPijt 0.527*** 0.567*** 0.592*** 0.621*** 0.589*** 0.603***
(4.55) (4.70) (4.67) (4.65) (6.62) (6.33)
LnExchangrateijt 2.881***
(2.70)
2.651**
(2.53)
2.373**
(2.13)
2.068*
(1.88)
1.881**
(2.03)
1.637*
(1.75)
ASEAN 2.520*** 4.040*** 3.278*** 6.943*** 2.171*** 3.767***
(6.09) (9.94) (6.31) (11.22) (4.45) (7.62)
Landlockedj 0.274 0.0517 -0.102 -0.392 0.123 -0.0406
(1.07) (0.20) (-0.34) (-1.29) (0.56) (-0.18)
Borderij 3.972*** 4.264*** 5.227*** 6.338*** 2.610*** 3.090***
(8.58) (9.98) (6.77) (10.04) (5.36) (5.11)
Colonyi 0.975*** 0.917*** 0.933*** 0.722*** 0.584*** 0.478***
(6.83) (5.29) (4.72) (3.51) (4.42) (3.13)
LnDistance 0.945*** 1.158*** 0.553***
(4.98) (4.81) (3.09)
1<km≤4000 2.844*** 5.167*** 2.449***
(4.73) (6.41) (3.19)
4000<km≤7800 2.687*** 5.438*** 2.393***
(4.48) (6.89) (3.04)
7800<km≤14000 3.330*** 6.193*** 2.804***
(5.65) (8.03) (3.64)
14000<km 4.117*** 6.953*** 3.072***
(6.67) (8.22) (3.91)
Constant -22.12*** -16.83*** -33.93*** -29.99*** -33.31*** -31.13***
(-9.79) (-10.14) (-12.02) (-15.93) (-16.83) (-20.61)
Observations 615 615 614 614 615 615
Adjusted R2 0.66 0.66 0.61 0.61 0.68 0.68
Robust t-statistics in absolute value in parentheses, * significant at 10%, ** significant at 5%, *** significant at 1%. OLS regression with year effects
88 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Table 13. Gravity factors and export margins (rice firms)
(1) (2) (3) (4) (5) (6)
Independent
variables
Total exports
(ln)
Total exports
(ln)
Total quantity
(ln)
Total quantity
(ln)
Total number of firms
(ln)
Total number of firms
(ln)
LnGDPjt 0.407*** 0.403*** 0.374*** 0.370*** 0.381*** 0.400***
(6.10) (6.22) (5.25) (5.26) (6.87) (7.02)
LnGDPjt/LBjt -0.952*** -0.883*** -1.107*** -1.046*** -0.564*** -0.462***
(-4.91) (-4.84) (-5.55) (-5.57) (-3.73) (-3.54)
LnDGDPijt 0.0270 -0.0383 0.135 0.0749 0.0572 -0.0409
(0.22) (-0.35) (1.08) (0.65) (0.67) (-0.64)
LnExchangrateijt 0.361
(0.29)
-0.789
(-0.69)
0.670
(0.48)
-0.490
(-0.39)
1.696
(1.53)
0.714
(0.75)
ASEAN 2.722*** 0.617 3.023*** 0.823 0.908 0.411
(4.29) (0.53) (4.76) (0.77) (1.63) (0.36)
Landlockedj -2.415*** -2.453*** -2.494*** -2.540*** -2.316*** -2.289***
(-6.15) (-6.30) (-7.19) (-7.18) (-9.41) (-9.55)
Borderij -2.100** -3.194*** -2.047** -3.212*** -0.299 -0.538
(-2.36) (-2.72) (-2.36) (-2.98) (-0.40) (-0.48)
Colonyi -0.341 -0.639 -0.355 -0.652 0.367 -0.0108
(-0.84) (-1.54) (-0.79) (-1.42) (1.41) (-0.04)
LnDistance -0.001 0.0491 -0.190
(-0.00) (0.16) (-0.72)
1<km≤4000 -3.039** -3.134** -1.456
(-2.12) (-2.27) (-1.08)
4000<km≤7800 -2.182 -2.209* -0.804
(-1.63) (-1.75) (-0.62)
7800<km≤14000 -1.974 -1.994 -0.483
(-1.50) (-1.60) (-0.38)
14000<km -4.672*** -4.688*** -2.608**
(-3.41) (-3.61) (-1.99)
Constant 20.47*** 22.74*** 14.66*** 17.44*** -0.226 -1.779
(6.24) (11.03) (4.39) (8.25) (-0.08) (-1.00)
Observations 487 487 487 487 487 487
Adjusted R2 0.23 0.29 0.21 0.27 0.23 0.29
Robust t-statistics in absolute value in parentheses, * significant at 10%, ** significant at 5%, *** significant at 1%. OLS regression with year effect
89 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Table 14. Gravity factors and export margins (wood and wood products firms)
(1) (2) (3) (4) (5) (6)
Independent
variables
Total exports
(ln)
Total exports
(ln)
Total quantity
(ln)
Total quantiy
(ln)
Total number of firms
(ln)
Total number of firms
(ln)
LnGDPjt 0.884*** 0.904*** 0.885*** 0.905*** 1.015*** 1.031***
(16.19) (16.21) (15.93) (15.70) (22.39) (22.32)
LnGDPjt/LBjt 0.0276 0.102 0.0914 0.142 0.314** 0.392**
(0.14) (0.53) (0.45) (0.70) (1.97) (2.49)
LnDGDPijt 0.644*** 0.571*** 0.578*** 0.542*** 0.790*** 0.710***
(5.00) (4.45) (4.51) (4.23) (7.45) (6.69)
LnExchangrateijt 3.036***
(4.31)
2.476***
(3.56)
3.077***
(3.71)
2.670***
(3.27)
3.434***
(4.45)
2.550***
(3.39)
ASEAN . 2.236*** . 1.599** . 3.746***
. (3.23) . (2.10) . (9.99)
Landlockedj -0.637** -0.695*** -1.029*** -1.121*** -0.454* -0.663***
(-2.29) (-2.61) (-3.43) (-3.83) (-1.72) (-2.76)
Borderij . 4.703*** . 5.432*** . 5.014***
. (12.06) . (9.51) . (11.85)
Colonyi 1.431*** 1.169*** 1.196*** 1.000*** 1.918*** 1.546***
(7.80) (6.51) (4.82) (4.04) (9.16) (7.47)
LnDistance -0.710*** -0.998*** -0.728***
(-5.52) (-5.99) (-5.80)
1<km≤4000 2.545*** 2.256** 3.288***
(3.14) (2.40) (5.48)
4000<km≤7800 1.368 0.574 1.839***
(1.64) (0.61) (2.95)
7800<km≤14000 1.682** 0.702 2.387***
(2.08) (0.75) (4.00)
14000<km 0.816 -0.124 1.240**
(0.97) (-0.13) (1.99)
Constant -0.671 -9.218*** -6.041*** -16.38*** -24.15*** -33.29***
(-0.35) (-5.33) (-2.72) (-9.15) (-14.51) (-25.22)
Observations 574 618 574 618 575 619
Adjusted R2 0.66 0.68 0.63 0.66 0.74 0.76
Robust t-statistics in absolute value in parentheses, * significant at 10%, ** significant at 5%, *** significant at 1%. OLS regression with year effect
90 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
countries in chapter 4, in this chapter, we find that this indicator positively affects firms’ margins for the
footwear and wood and wood products sectors whereas for the rice sector, the effect is negative but
insignificant. This finding indicates that Vietnamese footwear and wood and wood products firms are
still able to benefit from their low-cost of labor for exports to the international market.
The multilateral resistance factor affects positively and significantly all firms’ export margins in
footwear and wood and wood products sectors. However, for the rice sector, no significant effects are
found in three margins. Those positive effects indicate that when the Vietnamese currency is devaluated
against its importing country’s currency, export margins of Vietnamese firms in these two sectors are
encouraged.
There is still a significant positive effect from ASEAN dummy variable for all three sectors. The role of
regional economic integration remains important in that tax exemption and elimination continue to play
a vital role in facilitating trade between member countries. Surprisingly, the fact a country is landlocked
has negative and significant effect on the three margins of rice and wood and wood products sectors
shows that this factor remains a trade barrier for the margins of certain export sectors of Vietnam.
Moreover, firms of footwear and wood and wood products sectors only benefit from countries which
share a common border with Vietnam as well as from a colony country such as France. Rice firms do
not consider contiguous countries attractive which absorb large amount of rice import from Vietnam.
The regression result for geographical distance varies according to the particular export sector. This
indicator is considered to be a trade barrier since distant countries result in a higher transport cost for
international trade. A negative sign for geographical distance is also found in almost all gravity model
studies such as Papazoglou (2007), Carrère (2006) and Fitzsimons et al. (1999). However, a few studies
on gravity model at firm-level data such as Bastos and Silva (2010) found a positive effect of the
distance variable concluding that more productive firms were able to reach distant markets. In our study,
we find that the effect of geographical distance depends on different margins and sectors. In fact, the
footwear sector intuitively shows the positive sign of distance implying that it is able to overcome the
distance challenge. However, in the rice sector, for geographical distance, the regression gives both
positive and negative sign, which are not significant. Meanwhile, for firms of the wood and wood
products sector, it has a negative and significant effect on the value of exports and the number of firms
margins and negative effect on the quantity indicating that only a few sufficiently competitive firms can
enter distant markets.
5.7. Empirical results and discussion
5.7.1. Export value and the gravity model
Under the proposed gravity model, we continue to regress the variables mentioned in the previous
section on the export value and volume of firms. The reason why we use firm-country-year regressions
with the gravity factors is to avoid the problem of aggregation bias. Here, the export value at firm level
is decomposed based on the total export value of a particular export sector. As such, the regression
results are differentiated by export value and volume of firm-country clusters. Since a firm differentiates
91 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
its role in different markets and through periods of time, we treat a firm exporting to a particular country
as one group and the year is as the time dimension therefore we can observe these dynamic changes by
firms for different markets at a certain time. This provides a panel of identified firm-country and time t.
We also test for the year effect by adding four-year dummy variables to the regression models.
For comparison purposes, the equation has been tested for the OLS, fixed and random effects as shown
in Table 15. The OLS regression method is not helpful in discriminating an individual firm's export
activities across the destination markets thus causing bias. From observation, almost all point estimates
in fixed effects regression are considerably larger than those obtained from the OLS regressions. More
importantly, the adjusted R-squared obtained from the OLS approach is consistently low while
comparatively, fixed effect regressions give a higher level of adjusted R-squared thereby confirming the
importance of heterogeneous firms across destination markets. We also ran the random effects model in
which comparing exporting firm groups may cause a contradictory conclusion. We also generate five
categories for the variable distance in order to identify how differently transport costs affect firms in a
particular sector. Furthermore, the Hausman test reaffirms the fixed effects model as an appropriate
option.
For the market size effect, our empirical findings are in accordance with almost all previous studies on
the gravity model using country-level data (Feenstra et al. 2001; Glick and Rose, 2002; Kangas and
Niskanen, 2003; Roberts, 2004) and firm, product level data (Muuls and Pisu, 2009; Bastos and Silva,
2010; Crozet and Koenig, 2010; Nguyen, 2009). Specifically, high-income destination markets are
stimulus factor for exports of footwear, rice and wood and wood products firms but this coefficient is
not significant for rice and wood and wood products firms showing that there is insufficient evidence to
conclude the positive effect of the importing countries' GDP on rice and wood and wood export value of
Vietnamese firms. In contrast, we detect that the wealthy destination markets encourage export value of
footwear firms but discourage export value of rice and wood and wood products firms. It also should be
noted that there is not enough evidence to conclude the effects of the difference in Vietnamese GDP per
capita and its importing countries although this indicator gives negative signs for all three firms sectors.
It should also be noted that in the OLS regression models, the exchange rate between Vietnam and its
importing partner encourages export margins of footwear and wood and wood products firms, however
in the fixed effects models of the three sectors, insufficient evidence is found to conclude for the effects
of this indicator.
Since fixed effects model eliminates the time-invariant variables, we find no effect of such dummy
variables as "ASEAN" and "landlocked", “border”, “colony” as well as "distance". We derived them
from the random regression results and observed that "geographical distance" positively affects
footwear and wood and wood products firms' exports. In most studies using trade data by country,
distance is considered as a trade barrier since it increases the trade cost between country pairs.
However, distance does not necessarily reduce firms' exports even Bastos and Silva (2010) proved that
highly productive firms are still able to reach remote markets.
92 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Table 15. Export value of firms and gravity model
Independent
variables
Footwear firms Rice firms Wood and wood products firms
OLS
(1)
FE
(2)
RE
(3)
RE*
(4)
OLS
(5)
FE
(6)
RE
(7)
RE*
(8)
OLS
(9)
FE
(10)
RE
(11)
RE*
(12)
LnGDPjt 0.33*** 1.93*** 0.24*** 0.26*** 0.05** 0.25 0.03 0.01 0.19*** 0.27 0.11*** 0.11***
(27.34) (5.27) (13.90) (14.18) (2.36) (0.21) (1.45) (0.12) (22.19) (0.64) (10.81) (9.64)
LnGDPjt/LBjt 0.17** 0.84** 0.20** 0.27*** -0.47*** -0.02 -0.53*** -0.4*** 0.06 -0.05 0.01 -0.03
(2.47) (2.10) (2.45) (3.16) (-8.19) (-0.03) (-7.74) (-6.72) (0.87) (-0.14) (0.13) (-0.52)
LnDGDPijt 0.10** -0.23 0.07 0.02 -0.19*** -0.03 -0.13*** -0.1*** 0.09** 0.09 0.10** 0.11**
(2.39) (-0.79) (1.25) (0.41) (-4.75) (-0.51) (-2.92) (-3.14) (2.04) (1.13) (2.13) (2.35)
LnExchangrateijt 2.0***
(9.32)
-0.07
(-0.40)
0.12
(0.72)
0.05
(0.34)
-2.08***
(-4.95)
-0.18
(-0.30)
-1.49***
(-3.54)
-1.7***
(-3.98)
0.6***
(4.83)
-0.22
(-1.35)
0.22
(1.61)
0.13
(0.97)
ASEAN 0.39*** . 0.33*** 1.60*** 1.45*** . 1.11*** -0.98 -0.5*** . -0.63*** -0.22
(5.13) . (2.82) (5.14) (10.24) . (6.05) (-1.60) (-8.96) . (-8.59) (-1.35)
Landlockedj -0.01 . 0.05 -0.05 -1.69*** . -1.41*** -1.1*** -0.2*** . -0.17** -0.29***
(-0.25) . (0.72) (-0.65) (-8.71) . (-6.96) (-5.32) (-3.31) . (-2.38) (-3.89)
Borderij 0.44*** . 0.28 1.24*** -1.45*** . -1.27*** -2.8*** 0.65*** . 0.68*** 1.03***
(3.35) . (1.61) (4.96) (-5.61) . (-4.38) (-5.03) (8.15) . (7.33) (6.73)
Colonyi 0.45*** . 0.24 0.09 -1.05** . -1.29** -1.30** -0.2*** . -0.13 -0.19**
(4.54) . (1.36) (0.52) (-2.56) . (-2.33) (-2.34) (-3.51) . (-1.60) (-2.32)
LnDistance 0.40*** . 0.40*** -0.05 . -0.09 0.14*** . 0.09***
(13.17) . (8.40) (-0.75) . (-1.10) (5.52) . (2.90)
1<km≤4000 1.55*** -2.3*** 0.60***
(4.67) (-3.44) (3.50)
4000<km≤7800 1.50*** -2.3*** 0.45**
(4.51) (-3.71) (2.51)
7800<km≤14000 2.18*** -2.3*** 0.72***
(6.61) (-3.60) (4.17)
14000<km 2.73*** -3.3*** 0.50***
(8.19) (-4.74) (2.71)
Observations 23237 3961 35203
No firm-country 9220 2535 19032
R2
0.09 0.86 0.20 0.90 0.06 0.87
Ftest (p-value) 163.3*** 90.9*** 73.8*** 6.5*** 167*** 204***
Wald2
16 (p-value)
Hausman test
1168.7***
165***
1143.1** 604.9***
30.77***
650.9**
133***
2037*** 2088***
Notes: Robust t-statistics in absolute value in parentheses, * significant at 10%, ** significant at 5%, *** significant at 1%.
93 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
We are continuing to examine the effect of the ASEAN dummy on its member countries' trade since
some papers have asserted that ASEAN has lost ground in stimulating intra-trade because that many
countries have benefited from being fully integrated into the global economy (Nguyen, 2010; Nguyen
and Heo, 2009) thus decreasing their active participation in this smaller economic bloc. Surprisingly, the
ASEAN effect is somewhat paradoxical since on the one hand it acts as a stimulus for exports of
footwear and rice firms, while on the other, exports of wood and wood products firms are discouraged.
It is therefore clear that, ASEAN is no longer supportive factor for wood and wood products firms.
The question of whether landlocked countries constitute a barrier to foreign exporting firms is also
interesting. Basically, a landlocked country can not be a supportive factor because freight costs are
inevitably higher. Although negative signs are found for landlocked coefficients, the effect is only
significant for rice and wood and wood products firms. As such, the negative sign of landlocked dummy
in the regressions for those firms confirms the importance of maritime transport in the international
trade. In particular, with heavy bulk goods such as rice and wood and wood products, the use of bulk
carriers helps firms considerably by increasing transport capacity and keeping down costs.
Comparing the effects of “border” across sectors, on firms in the footwear sector find contiguous
countries attractive whereas rice and wood and wood products firms do not benefit from such a factor.
Vietnam’s status as a former colony does not result in greater export value in the in the rice and wood
and wood product sectors which runs counter to some studies using country trade level data where such
status is found to be consistently positive. In particular, the effect of “distance” is significantly positive
for the export value of firms active in footwear and wood and wood products sectors showing that at
firm level exports, geographical distance does not cause serious problems for those firms. We also
include five categories for the distance variable and find that for the footwear and wood and wood
products firms, all distance categories are consistently positive whereas they are all negative and
significant for the rice firms.
5.7.2. Export volume of firms and the gravity model
We also test the gravity model using firms' export volume in order to find how it is affected by different
factors of the gravity model. For example, heavy and bulky goods such as rice and wood and wood
products force exporters to pay higher freight costs when shipping to distant markets. For comparison
purposes, we regressed three types of model including the OLS, fixed effect and random effect models.
The Hausman test shows that the fixed effect model is the most appropriate option, which confirms the
role of firms in determining their export pattern. Concerning coefficients of fixed effect model in
column (2), (6) and (10) of Table 16 for footwear, rice and wood and wood products sectors
respectively, the signs of the time-variant variables are generally the same as those of export value
regression except for landlocked and border variables. However, rice and wood and wood products
firms show no significant effects from the fixed effect regression model therefore we do not have
sufficient evidence to confirm the effects of gravity factors on the export volume of these firms. The
GDP of the importing country positively affects footwear firms' export volume. The magnitude of the
coefficient is considerably higher in that for every 1% increase in the GDP of the importing
94 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
countries, firms' export volume increase by 2.5%. The GDP per labor force effect is positive confirming
the supportive factor for the export volume of footwear firms. The difference in the GDP per capita of
Vietnam and its importing countries gives mixed signs but insignificant for firms’ export volume in all
three sectors.
We also examine the effect of the time-invariant variables from random effect models for each specific
sector in the column (3), (7), and (11). For footwear and wood and wood products firms, we find that
distant markets affect positively on export volume implying that for those firms, higher transport costs
or other fees related to shipping these goods such as export/import duties, customs fees and port charges
do not matter. In contrast, distant markets are a larger challenge for rice firms therefore large quantity of
rice products are not exported to such markets. This may be due to rice being more likely to be a heavy
and bulky leading to relatively higher freight costs.
Surprisingly, ASEAN dummy effects appear to be inconsistent between sectors. While they give a
positive sign for the footwear and rice export volume but a negative sign for wood and wood products
export volume. Indeed, a smaller quantity of wood and wood products is exported to ASEAN countries
compared with a larger quantity of footwear and rice products exports. However, although the
landlocked dummy is still negative for all three sectors, it is only significant for the rice and wood and
wood products sectors indicating that landlocked countries are consistently at a disadvantage
particularly for these sectors. Notably, the magnitudes of all landlocked coefficients for firms’ export
volume with gravity factors are larger than those of export value indicating that the landlocked effects
on firms’ export volume are more sensitive than on their export value.
95 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Table 16. Export volume of firms and gravity model
Independent
variables
Footwear firms Rice firms Wood and wood products firms
OLS
(1)
FE
(2)
RE
(3)
RE*
(4)
OLS
(5)
FE
(6)
RE
(7)
RE*
(8)
OLS
(9)
FE
(10)
RE
(11)
RE*
(12)
LnGDPjt 0.36*** 2.52*** 0.29*** 0.35*** 0.03 -0.35 0.02 -0.01 0.22*** -0.40 0.14*** 0.15***
(28.35) (5.61) (16.59) (18.61) (1.39) (-0.25) (0.78) (-0.20) (22.78) (-0.77) (11.76) (11.20)
LnGDPjt/LBjt 0.22*** 0.97** 0.22** 0.24** -0.56*** 0.54 -0.55*** -0.4*** 0.05 -0.24 0.02 -0.01
(2.89) (1.98) (2.44) (2.54) (-8.20) (0.43) (-7.43) (-6.37) (0.61) (-0.49) (0.25) (-0.12)
LnDGDPijt 0.15*** -0.39 0.12** 0.08 -0.11** -0.07 -0.10** -0.11** 0.03 0.04 0.01 0.03
(3.17) (-1.05) (2.10) (1.45) (-2.32) (-1.16) (-2.00) (-2.30) (0.69) (0.39) (0.37) (0.69)
LnExchangratejt 2.0***
(9.01)
-0.05
(-0.27)
0.29
(1.48)
0.25
(1.02)
-1.33***
(-2.74)
-0.01
(-0.02)
-0.874*
(-1.86)
-1.2***
(-2.60)
0.9***
(5.93)
0.13
(0.66)
0.66***
(4.24)
0.63***
(4.04)
ASEAN 0.64*** . 0.72*** 3.48*** 1.60*** . 1.31*** -0.48 -0.15** . -0.26*** -0.45***
(8.14) . (6.25) (11.60) (11.08) . (7.20) (-1.10) (-2.29) . (-3.34) (2.76)
Landlockedj -0.09 . -0.07 -0.20** -1.64*** . -1.44*** -1.2*** -0.4*** . -0.37*** -0.43***
(-1.52) . (-0.84) (-2.27) (-9.56) . (-7.88) (-6.34) (-6.45) . (-5.00) (-5.67)
Borderij 1.15*** . 1.07*** 3.08*** -1.09*** . -0.92*** -2.2*** 1.15*** . 1.11*** 1.76***
(8.27) . (5.86) (12.19) (-4.71) . (-3.69) (-5.73) (12.61) . (10.68) (11.58)
Colonyi 0.54*** . 0.44*** 0.23 -0.84** . -0.91** -1.01** -0.3*** . -0.28*** -0.30***
(5.81) . (2.92) (1.53) (-2.42) . (-2.14) (-2.34) (-5.87) . (-3.35) (-3.48)
lnDistance 0.43*** . 0.43*** -0.01 . -0.04 0.11*** . 0.07**
(13.74) . (9.08) (-0.09) . (-0.46) (3.94) . (2.14)
1<km≤4000 3.14*** -2.1*** 1.02***
(9.96) (-4.00) (5.69)
4000<km≤7800 3.27*** -2.1*** 0.84***
(10.24) (-4.26) (4.59)
7800<km≤14000 3.91*** -1.8*** 0.96***
(12.42) (-3.87) (5.44)
14000<km 4.35*** -3.0*** 0.89***
(13.55) (-5.35) (4.54)
Observations 23106 3931 34394
No firm-country 9179 2513 18628
R2
0.10 0.82 0.18 0.88 0.06 0.85
Ftest (p-value) 186*** 53.9*** 74.4*** 1.81* 182*** 211***
Wald2
13 (p-
value)
Hausman test
1447***
156.82***
1439*** 548.83***
18.1**
580***
127***
2274***
2330***
Notes: Robust t-statistics in absolute value in parentheses, * significant at 10%, ** significant at 5%, *** significant at 1%.
96 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
5.8. Conclusion
The chapter aims to answer the question of whether firms' exports are differently affected by gravity
factors compared with the total exports of a country. Our study differs from the previous ones such as
Bastos and Silva (2010), Hillberry and Hummels (2008) in examining gravity factors for three separate
export sectors. We observe that different effects are obtained corresponding to each determinant
suggesting that not only firm but also sector characteristics play an important role in determining the
pattern of the Vietnamese exports.
Based on the Hausman test, we find that fixed effect model is the most appropriate option which means
that each firms' individual characteristics cause the changes in the gravity export determinants as well as
the characteristics of the exporting countries. However, a disadvantage of using the fixed effect model is
that it is not possible to obtain coefficients of the time-invariant variables since these variables are
removed during regression.
The signs of the importing countries' GDP corresponding to footwear, rice and wood and wood products
exports are in line with the one obtained from country level data confirming the stimulus role of the
economic size of the importing countries. Meanwhile, the GDP per labor factor has a negative sign for
rice and wood and wood products firms sectors but this is not significant for wood and wood products
firms. Generally, the difference in GDP per capita between Vietnam and its importing countries
negatively affects the footwear and rice firms' export value but this coefficient is not significant.
Using random effect regressions, geographical distance between Vietnam and its importing countries
shows a positive and significant sign for footwear and wood and wood products firms indicating that
transport costs do not matter and discourage firms to reach distant markets. Although this indicator is
not always negative, nevertheless rice firms find it a challenging factor in international trade. Although,
ASEAN role in the regional economic integration remains a generally supportive element for footwear
and rice exporting firms, it has a discouraging effect on the export activity of wood and wood products
firms.
In short, examining the role of firms in determining their trade pattern is another fascinating story, as
firms are heterogeneous and dynamic to varying degrees depending on their destination markets.
Normally, aggregate export value of a country has some bias because firms may face fierce competition
in different markets and therefore need to become more active and dynamic to succeed. However, if we
also decompose the aggregate export value of country into firms' extensive and intensive margins and
other authors' related works, the regression results may not reflect the role of firms in a particular sector.
Although our work, to some extent has resolved the latter problem, we were not able to capture the
gravity effects for all export sectors for the reasons given earlier.
97 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
CHAPTER 6. DETERMINANTS OF FIRMS' EXPORTS: A CASE OF A DEVELOPING AND
EMERGING COUNTRY
SUMMARY
This chapter examines Vietnamese firms’ internal key factors affecting their export performance
by using a 2008 data set of firms in three export sectors, footwear, rice and wood and wood
products. We transform the dependent variable, which lies strictly in the unit interval, into logit
and apply the OLS regression method. We also attempt to carry out quantile estimation with
bootstrapped standard errors, because we are concerned about the severe extreme values of firms'
export intensity and non-normality of the residual. A significant positive sign of firm size in
footwear and wood and wood products sectors continues to confirm the advantage of economies
of scale. In contrast, firm size decreases the export intensity of the rice firms at 10th quantile,
indicating that economies of scale do not encourage these firms' propensity to export. In
particular, human and capital intensity are negatively associated with export intensity indicating
that highly skilled workers are not needed and firms restrict their fixed assets investment in these
low-technology manufacturing firms. State-owned rice and wood and wood products firms are
found to be unsupportive to export intensity, though for the footwear sector, SOEs encourage
their propensity to export at 25th quantile. Moreover, firms' longer presence does not confirm
their better export performance, with the greatest level of significance shown across quantiles for
the wood and wood products sector.
98 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
6.1. Introduction
The important role of exporting firms in international trade has been asserted in many studies and such
smaller economic entities do not interact in the same way as their countries. How firms operate in
foreign markets depends not only on their own ability but also on their environment as well as when
firms enter the market or whether they continue working effectively or exit after a period of time. One
important strand in international trade literature focuses on the dynamic of firms, which helps them be
more competitive with their foreign partners. Bernard et al. (2007ab) have emphasized the importance
of exporters capable of growth becoming more productive, skill- and capital-intensive than non-
exporting firms.
In Vietnam, exports have made a large contribution to the annual national income. In recent years, its
revenue has risen dramatically and in 2012 peaked at 114 billion US dollars, an increase of 18.2%
compared with 2011. The top of the list of export products are crude oil, textiles and garments,
footwear, electronics, wood and wood products, rice and machinery. Vietnam's main export markets
include the United States of America, Japan, China, Australia, Singapore, Germany and the United
Kingdom.
Despite the major contribution of Vietnam's exports to the country's income, there have been relatively
few studies examining determinants on firms' export performance. Nguyen and Hiroshi (2008) addressed
the question of whether the self-selection and learning-by-doing hypothesis is applicable to the exporting
firms of Vietnam by including several important factors such as TFP, labor productivity, age of firm,
type of firm ownership, etc. Although a random sample of 1,150 firms was collected based on the
population of 0.21 million Vietnam's manufacturing firms, their findings were reliable since no selection
bias was found for Heckman's random-effects dynamic probit model. Their paper also proved the
exceptional performance of Vietnam's exporting firms compared with non-exporting firms. Vu (2012)
attempted to examine whether firms' productivity can explain their decision to export by using a
balanced dataset from 2005-2009. However, Nguyen and Hiroshi (2008) unlike Vu (2012) showed that
the learning by exporting hypothesis is invalid for Vietnam's private manufacturing firms. The authors
also found that export status of firms had no significant effect on TFP growth, change of scale, technical
efficiency and technical progress. In a developing country such as Vietnam, export-led growth strategy
and trade liberalization play a crucial role in developing the country's trade and firms' activities. By
using firm-level data between 2000 and 2009, Doan and Kiyota (2014) were able to prove that although
trade liberalization encouraged aggregate productivity growth in Vietnamese manufacturing, it did not
necessarily have negative effects on small firms which achieve even better opportunities by being more
productive.
This chapter contributes to the existing literature of Vietnam’s export trade in that it examined internal
important determinants of exporting firms active in three sectors such as footwear, rice and wood and
wood products which are considered to be key exporting sectors of the country. The remainder of this
chapter is organized as follows: Section 6.2 presents related literature on firms' characteristics, section
6.3 demonstrates the location of footwear, rice and wood and wood products firms across regions of
99 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Vietnam, section 6.4 explains the methodology and the data, section 6.5 reports and discusses
determinants of these firms' export performance based on empirical results and section 6.6 concludes the
chapter.
6.2. Related literature
In recent years, many researchers have become increasingly concerned about micro level data since the
aggregate export data are biased, to some extent. One possible plausible explanation is that exporting
firms' performance in each industry reflects their different dynamics in international trade. Literature on
exporting firms' activity reflects considerable interest in many determinants, which can be classified into
two the main groups in Figure 22 below.
Studies on exporter's performance mainly focus on export revenue, export productivity, profitability and
export intensity (propensity to export). Moreover, determinants of firms' performance cover firms'
characteristics and activities. Those factors affect firms' export performance differently depending on the
type of data as well as the methodology applied to analyze the interaction between those factors and
export performance.
- Firm's age: The age of a firm denotes the year of its establishment. Firms who have a long history of
foreign market operations are more experienced than others since their exporting capabilities may have
been improved after some years cooperating with foreign partners. However, in an ever more fiercely
competitive international market, no guarantee for survival. Although experienced firms are frequently
thought to be more successful than newborn ones, this does not always mean that older firms work more
efficiently.
Figure 20. Firm's factors in international trade
Sources: Author's diagram
Firms' factors
Firms' performance Firms' characteristics
- Productivity
- Labor cost
- Capital intensity
- Innovation
- Training
- Learning by exporting
- Earning profitability
- Enter, exit and survival
- Age
- Size
- Foreign market knowledge
- Educated workers
- Workers' gender
- Workers' skill level
- Location
- Affiliation
- Export assistance
- Ownership
100 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
There are numerous studies investigating how firms' performance is affected by firm age. One stream of
research on firm age showing evidence of its positive impact on their export performance includes
typical works by Majumdar (1997), Iyer (2010), Fakih and Ghazalian (2013), Javalgi et al. (2000).
According to Javalgi et al. (2000) the aggregate level data, the age of firms confirms their export
intensity although the relationship for manufacturing-based service providers is not significant. To be
discussed later in this chapter, these Indian older firms are found to be more productive but less
profitable than the younger ones.
Another research stream has shown that the age of firms is not necessarily a positive factor on
international markets. In fact, some older firms are not as flexible in adjusting to dynamic foreign
markets whereas young firms are better able to adapt to market needs of service sectors. Interestingly,
Caparas (2006) investigated the seemingly paradoxical relationship between firm age and export sales.
He found that while older Philippine firms in the clothing and food manufacturing sector are successful
in exports because of longer time they have had to learn about the market nevertheless younger
electronic firms are able to penetrate effectively the export markets due to the fast pace of change in the
technology sector.
- Firm size: The size of a firm reflects the notion of economies of scale, which in theory induces the
growth in export turnover. Most studies use the number of workers as a good proxy for firm size such as
Bonaccorsi (1992), Bleaney and Wakelin (2002), Wakelin (1997), Muûls and Pisu (2009), Farole and
Winkle (2011), Javalgi et al. (2000), Aggrey et al. (2010), Samiee and Walters (1990), Roper and Love
(2002). In other cases, firms' revenue can be a proxy for the size of firm as shown by the studies of
Papadogonas et al. (2007), Greenway et al. (2004). Nevertheless, Obben and Magagula (2003), Javalgi
et al. (2000) used for comparison purposes the number of employees and firms' revenue as proxies for
firm size in their studies.
The firm size advantage for export performance is a widely accepted as fact of modern industrial
economies. A persuasive explanation is that firms can take advantage of factor endowment in which the
labor factor is considered as the economies of scale and firm size increases significantly with exports
(Javalgi et al. 2000; Farole and Winkler, 2011). However, the studies by Wagner (1995) and Robson et
al. (2012) reveal the positive but decreasing relationship between firm size and export intensity.
Specifically, the impact of firm's size is not the same across sectors and size categories.
- Firm's average wage: This indicator is a proxy for the labor cost so called human capital intensity.
Literature on exporting firms' performance shows a mixed correlation between their labor cost and their
export intensity (or propensity to export). One strand of the literature in favor of using firm's average
wage as indicator explains that certain firms do not require skilled and highly educated workers
(Papadogonas et al. 2007). Low labor cost can be seen as a competitive advantage for Greece, which
tends to export low-tech products such as food products, beverages and tobacco, textile and footwear
products, wood and cork products. The negative sign of firm labor cost for export performance is also
confirmed by Liu and Shu (2003), Bhavani and Tendulka (2001), Michiel (2002) where manufacturing
industries have an international labor cost competitive advantage.
101 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
However, low labor cost is not an advantage for most of the developed economies because firms need to
pay higher wages in order to increase export turnover. Wagner (1995) found a positive effect of the
human capital intensity on export intensity for German manufacturing firms in four sectors: basic
products, capital goods, consumer goods, food, beverages and tobacco, although for basic goods, the
effect is not important. Another study by Greenway et al. (2004) confirms such positive relationship
between the labor cost of British domestic-owned firms and their export decisions and propensity to
export. One important feature of this study is that exporting firms also produce for the domestic market
therefore it is debatable how the presence of MNEs affects the export behavior of other domestic
manufacturing firms. A relevant explanation for the positive sign of labor cost is that highly skilled labor
is very important in the context of the highly competitive labor market in European countries such as the
UK.
Nonetheless, production costs do matter when large firms plan to expand their market shares by
reaching out to distant markets as part of their strategy to extent product life cycle. When large
enterprises such as the multinational national enterprises are seeking new markets, labor cost is of
crucial importance. American MNEs are typical of export-oriented investors with affiliates located in
less developed countries such as Taiwan, Hong Kong, Malaysia, and Thailand. Those firms therefore
employ low-wage workers in the local markets to increase export revenue.
Firm's capital intensity: This indicator measures a firm's efficiency in deploying its assets and indicates
how much capital investment is needed in order to yield a single dollar of sales revenue. It also indicates
a firms' level of production technology. In most cases, capital intensity is conducive to a firm's export
performance thus providing a positive sign (Özçelic and Taymaz, 2003; Wagner, 1995; Wakeline, 1997
and Rankin et al. 2006). Interestingly, Özçelic and Taymaz (2003) have found a significant and positive
impact of this factor on both innovators and non-innovators' export intensity for Turkish manufacturing
firms active in nine export sectors35
. In the case of the Turkish economy considered to be a labor-
abundant country, this finding is reminiscent of the Leontief paradox.
Nonetheless, the effect of capital intensity is somewhat different in the case of German manufacturing
firms (Wagner, 1995). While the positive sign of capital intensity explain the export intensity of capital
goods, consumer goods, food, beverages and tobacco, the reverse result is found in basic goods
industries. Likewise, Bleaney and Wakelin (2012) investigated the negative impact of capital intensity
on export intensity for British manufacturing firms and arrived at a 10% significance level.
Firm innovation: There is a growing literature on the role of innovation in international trade. In fact,
technological firms devote a large amount of capital to R&D activities in order to modernize equipment
and machinery. There are a number of studies exploring the possible positive effect of R&D on export
performance: Roper and Love (2002), Sterlacchini (2001), Papadogonas et al. (2007), Lin and Shu
(2003), Robson et al. (2005), Wagner (1995), Wakelin (1997), Bleaney and Wakelin (2002), Roper et
35
Sample of Turkish exporting sectors covers Food, Textile, Wood, Paper, Chemicals, Non-Metallic, Metal, Engineering and
Other.
102 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
al. (2006), Banga (2006) and Michiel (2002). However, we note the exception for Canadian firms
(Lefebvre et al. 1998) where R&D intensity was found to be negative for such firms' export
performance in North American and global markets. For the case of Vietnam, innovation is also
considered to encourage a firm’ competitiveness by increasing productivity. Nguyen et al. (2009) have
used a database of Vietnamese exporting firms in the 2005 SME survey by the GSO and proved that the
probability of firms in export status increases with their innovation activities which include product
innovation, process innovation and product modification. For a developing country such as Vietnam,
since the country’s Doi Moi implementation, innovation presents as a breakthrough approach for
national economic development as well as for a firm’s competitiveness though patent grant (Vu, 2012).
Firm productivity: Literature on firms in the international trade reveals the stylized fact that productivity
raises exporting revenue for both manufacturing and service firms. A striking list of papers successfully
explained that highly productive firms could engage with and reach more distant markets (Bastos and
Silva, 2010; Bernard and Jensen, 2004ab; Bernard and Wagner, 1997; Bigsten et al. 2000; Clerides et al.
1998; Fernandes and Isgut, 2005; Wagner, 2007; Muûls and Pisu, 2009). By including firm's
productivity and export decision in a regression equation, Bernard and Jensen (2004a) clearly
demonstrated that the TFP is a main source of firm' success. Moreover, Bernard and Jensen (2004b)
show that productivity itself predates firm' entry into exporting. Wagner (2007) also highlighted that
export activity does not necessarily enhance labor productivity or total factor productivity. The question
of whether exporting in turn raises productivity is also addressed by Biesebroeck (2005), Loecker
(2007), Mukim (2011), Delgado et al. (2002).
Firm's ownership: The ownership of exporting firms is of crucial importance since the notion itself not
only specify the owner but also reflects the role of the management factor in a firm's capability to create
value both within its home country and in export markets. Three main types of firm's ownership: state-
owned firms, private and foreign-owned firms are most frequently mentioned in the literature on
exports. While foreign-owned firms emerge as the most competent type of enterprise in international
markets, state-owned firms tend to have a depressed export performance (Aggrey et al. 2010; Rankin et
al. 2005; Javalgi et al. 2000; Farole and Winkle, 2011; Özçelic and Taymaz, 2003).
However, the effect of firm's ownership needs to be seen in the context of its country of origin. In less
developed and developing countries where corruption in state-owned enterprises is increasingly
common, firm's business activities are badly affected. In contrast, the prominent characteristics of
foreign-owned firms are clearly shown through good management skills, technology transfer and
support in expanding their markets. Moreover, many foreign-owned firms have developed a worldwide
distribution network and outsourced to the less developed countries to benefit from low-cost labor
thereby creating competitive export products. Aggrey et al. (2010) found that foreign ownership was
positively correlated with exports from two Eastern African countries: Kenya, Tanzania and insignificant
negative sign for Ugandan exports. The study also confirms that foreign ownership is a vehicle for the
international transfer of management skills, technical knowhow and market information.
103 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Firm's geographical location: Firm location indicates where a firm is established and carries out its
business operation. For an exporting firm, location may be an advantage if near the source of raw
material, and high density of population. On the other hand, firms may also find it attractive to open
their establishments in large centers of population, which encompass places such as cities or towns
where convenient administrative structures are available. In some cases, geographical location relates to
firms' tendency to establish an industrial clusters since firms active in the same sector are more likely to
concentrate around other larger firms, which encourages the creation of a dense distribution and supply
network. In a regression equation, this indicator is treated as a category variable thus enabling comparisons
between categories (Aggrey et al. 2010; Farole and Winkler, 2011; Campa and Guillén, 1999; Freeman
et al. 2010; Zhao and Zou, 2002; Freeman et al. 2012). Zhao and Zou (2002) described the typical case
of China with its vast land mass and it has far more advanced coastal areas where firms' location clearly
matter. Specifically, the result shows that firm location positively affects the export intensity and
propensity of Chinese firms and those firms located in coastal areas, which have higher odds of
committing to exporting than firms located inland. Whereas, Freeman et al. (2012) have divided Australian
firm location into two categories metropolitan and regional areas. Although normally, firms in metropolitan
areas have an advantage over those in regional areas, no positive impact is found for their export
performance.
6.3. Location of Vietnamese footwear, rice and wood and wood products firms
Many studies assert that the role of international trade is to help create an effective and better utilization
of natural resources, which particularly contributes to the well-being of the people. Since Vietnam's
economic reform in 1986, this government policy has stimulated domestic enterprises to engage with
the international market by applying 0% tax to almost all export goods whereby Vietnamese exports
today make a major contribution to the country's annual income.
Market liberalization encourages access by firms from developed countries whose advanced technology
is applicable to most production sector. In the case of Vietnam, multinationals entered the market in
different ways. Firstly, foreign owned firms establish their own firms in the country. Secondly, they
conclude processing agreements with local firms, which include a processing fee and transfer of
technology, machinery, designs or raw materials. The presence of foreign owned firms also helps create
employments for the country which has grown substantially from 0.35 million workers in 2000 to 1,726
million in 201036
.
Vietnam has 3444km of coastline, is mostly hilly and densely forested with tropical forest cover of
around 42% and has low labor cost, all these characteristics give it a huge advantage for its major export
products which are fishery products, wood and wood products, rubber and many other agricultural
products such as rice, coffee, and cashew nuts. Of the annual major exports, articles of apparel and
clothing accessories dominate, earning 11.2 billion US dollars in 2010. In the same year, the footwear
sector ranked second in the list of the top 10 largest export products capturing 5.12 billion USD market
share with wood and wood products being the sixth largest export sector with a revenue of 3.4 billion
36
Statistical Year Book 2011 of Vietnam on Population and Employment issued by GSO
104 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Figure 21. Vietnam's number of firms across regions in 2008
Source: Author's diagram derived from the VCO data on firms' exports
105 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Figure 22. Vietnam's export value of firms across regions in 2008
Source: Author's diagrams derived from the VCO data on firms' exports
ffffirmfirm export
106 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
USD. Vietnam is also a second largest rice exporter after Thailand with an approximate revenue of
3.24 billion USD in 2010.
The number of exporting firms in wood and wood products, footwear and rice sectors varies between
the sixty-four provinces of Vietnam. In 2008, the four largest exporting provinces were Ho Chi Minh,
Binh Duong, Dong Nai, Ha Noi which account for 68.7% and 75.1% of the total number of firms active
in footwear and wood and wood products sectors whereas Ho Chi Minh, Can Tho and Ha Noi
provinces included the largest number of exporting rice firms. Meanwhile, in 2008, there were 1234
firms exporting wood and wood products, followed by 237 footwear-exporting firms and 83 rice-
exporting firms. (see Figure 21).
The five provinces of Binh Duong, Dong Nai, Ho Chi Minh, Binh Dinh and Ha Noi accounted for the
majority of the country exports of wood and wood products out of which the export value of Ho Chi
Minh province firms was 3,875.03 billion VND in 2008. Five largest footwear exporting provinces were
Ho Chi Minh, Dong Nai, Binh Duong, Hai Phong and Long An out of which firms in Ho Chi Minh city
accounted for 3,748.83 billion VND in the same year. The rice exporting firms of the South East region
(Ho Chi Minh), Mekong River Delta (Kien Giang, Can Tho and An Giang) and the Red River Delta
(Ha Noi) accounted for most of the country's rice exports out of which Ho Chi Minh city is also the
largest rice exporting province earning 1,835.68 billion VND for the same year. (see Figure 22).
6.4. Methodology and data
6.4.1. Description of variables
The chapter examines determinants with variables affecting firms' exports as follows:
Dependent variable: We focus on firms' export intensity which is constructed by dividing their export
value by total revenue. The revenue, export value, labor cost and net fixed asset in Vietnamese dong
(VND) are based on trade activities by an individual firm for the year 2008.
The average value of export to revenue ratios for wood and wood products firms is the largest, followed
by footwear and rice firms with the percentage of 29.6% and 21.6% respectively. The difference in
export intensity of firms sectors shows that wood and wood products firms seems to better
internationalize whereas rice firms remained mainly focused on the domestic market.
Independent variables:
Wage of employee: The wage of employee is calculated by dividing the labor cost by the total number
of employees and corresponds to the cost of the quality of labor.
Size of firm: Basically, the number of employees or the total annual revenue of a certain firm can be a
proxy for firm size. In this chapter, we used total employees to check whether the theory of economies
scale is relevant.
107 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Capital intensity: We establish the capital intensity by dividing the net fixed assets of a firm by the
number of employees. This may also provide information on the use of more advanced production
technology.
Firm ownership: In Vietnam, exporting firms may have different types of ownership37
. To express
simply, we have classified the dummy variable firm ownership into state-owned firms and non-state-
owned firms. In 2008, state-owned firms dominate other types of firms and account for almost 79% of
all rice firms. There are a large number of state-owned firms active in the wood and footwear sectors
with nearly 56% and 44% respectively.
Firm age: According to Shumway (2001), firm age is the number of years since a firm was first listed.
Nonetheless, we follow the definition of firm age that most of the studies on international firms define
as the number of years from the establishment of a legal entity. In general, although the average age of a
firm in the three sectors is relatively low, the range of firm ages is high. In fact, the age of the oldest
exporting firm is 57 and the youngest is 1 and firm ages ranging from one to nine years account for 75%
of the total, indicating that most of the firms are young compared with the majority of firms from
developed countries.
Firm location: Officially, eight locations known as socio-economic zones and approved by the
government of Vietnam include the following: Red River Delta, North Eastern, North Western, North
Central Coastal, South Central Coastal, Central Highlands, Mekong River Delta and South Eastern. For
the footwear sector, the majority of firms located in the South Eastern region account for almost 74% of
the total firms, followed by the Red River Delta where the number of footwear firms accounts for 17%
of the total. However, rice firms are evenly located throughout the regions especially the Red River
Delta, Mekong River Delta and South Eastern region, which include most of the rice firms which
account for almost 85% of the total for this sector. For the wood and wood products sector, the majority
of the firms are located in the South Eastern region, Red River Delta and South Central coastal showing
that such firms are not evenly distributed across the regions.
In fact, Vietnam's exporting firms of footwear, rice, wood and wood products sectors are concentrated
mainly in the central, northern and southern regions. The percentage of firms operating in the other sub-
regions mentioned above is so small that only three dummy variables for the location of firms have been
added in the model. Nevertheless, two firm location dummy variables are included to avoid a possible
problem of multicollinearity. Except for firm wages and total employees of firm, which are transformed
into natural logarithms, firms' capital intensity is left untransformed since its negative values are
significant. As such, our model is as follows:
2
0 1 2 3 4 5 6lnWAGE lnEMPLOYEE lnEMPLOYEE CAPINTENSITY lnFIRMAGE STATEOWNERi i i i i i
i
EXINT
LocationDummies
37
According to the law of Vietnam, firm types include State-owned enterprise, limited liability enterprise, joint-stock
enterprise, one member limited enterprise, partnerships, private enterprise, group of enterprises, foreign direct investment
firms
108 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
6.4.2. Regression method
As proposed by Papke and Wooldridge (1996), we implemented the logit transformation for the
response variable's values (export intensity) which strictly fall within the unit interval (see Appendix 3.1,
3.2, 3.3). Logit function helps transform a continuous variable (usually probability p) in the unit interval
into the real line where it is usually the logarithm of the odds therefore the transformed response
variable y* has the following form
* log( )1
yy X
y
We used the Ordinary Least Squared (OLS) approach to model y* as a linear function of a set of the
proposed regressors. The linear predictions of this regression, when fed through the inverse logit
transformation, take them back to the space of the original proportion variable, and constrain them to lie
within the unit interval.
One important point to note is, the export intensity values of firms in the three sectors largely fall
between (0, 0.2], accounting for around 50% of the total. Here we find that the logit transformation of
the response variable yields many negative values and the kernel density estimates of the transformed
variable seem not to follow normal distribution. This fact leads us to take account of a possible problem
arising from extreme values of both dependent and independent variables and the heterokendasticity.
Where the dataset contains a significant number of extreme values, the OLS method predicting the
linear model between a set of regressors and the outcome variable based on the conditional mean
function ( | )E y x may not yield efficient estimators. We are therefore interested in obtaining regression
coefficients for the relationship between the regressors and regressand by using the conditional median
function ( | ).qQ y x While estimators obtained from OLS based on minimizing the total sum of squared
error 2( ),ii the quantile regression method known as least-absolute-deviations (LAD) developed by
Honoré (1992) minimizes | | .ii More importantly, quantile regression is a better alternative method
if the errors are highly non-normal. This method, providing a richer characterization of the data, allows
us to entangle the impact of a covariate on the entire distribution of the response variable rather than its
conditional mean (Baum, 2013).
6.4.3. Data
6.4.3.1. Data source and description
Data on firms' exports are obtained from the VCO for the year 2008. For each export transaction at a
given point in time, firms begin the custom clearance procedure by providing all information required
on a declaration form. Specifically, a declaration sheet includes information on date of declaration, firm
identity code, importing country code, transaction code, currency code, exchange rate at time of
declaration, export volume, unit price, and export value expressed in VND.
109 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Export transaction requires a legal export contract signed by the Vietnamese firms and its importing
partner. The single customs declaration including the export contract allows a firm to make several
shipments of varying sizes to its partner for a specific period of time and for a specific product.
Although, the export value can in fact, be expressed in a number of major currencies such as USD, JPY,
EUR, AUD, VND, THB, SGD, the actual export value expressed in VND is used for convenience to
run the regression.
International trade transactions are based on international trade terms, which commonly include FOB
and CIF. According to Incoterms 2013 (International Trade Terms) issued by the ICC (International
Chamber of Commerce), firms who sign a DDP contract, maximize the exporter's obligation therefore
becoming more attractive for certain clients. In practice, Vietnamese firms mostly conclude FOB
contract for exports and CIF contract for imports, which are satisfactory balance between right and
obligation of two parties.
Data for the right-hand side regression model namely firms' revenue; firms' fixed assets, firms'
depreciation values or the number of firms' employees were provided by the GSO who conducts firm
survey in all Vietnamese regions and uses the following three main methods:
- Face to face interview: A survey assistant makes a direct interview with a director of firms then write
the answer in a questionnaire. This method is applied to firms who have not fully complied with
national accountancy regulation and therefore illegible to fill out the questionnaire either because they
are too small, about to be dissolved or have come under suspicion.
- Indirect survey method: The GSO organizes training for chief accountants of firms who then become
qualified to fill out a survey questionnaire.
- Web-based survey: This method is commonly applied to firms with internet access.
A survey questionnaire comprises 18 sheets requesting three main categories of firms' information:
* General information on firms:
- Name of firm;
- Address, telephone number, fax, email;
- Type of firm
- Economic activity
* Information on number of staff and their wages
- Number of workers
- Wages
* Information on production and firms' business activity
- Asset and capital source;
- Firms' revenue;
- Taxes, fees and charges paid to the State;
- Investment capital;
- Research and development investment capital;
110 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
- Firm's innovations and its effectiveness;
- Turnover of goods and services categorized into 168 product groups;
- Production costs categorized into materials and services originated from local or foreign country.
Although the survey questionnaire includes various information of enterprises in the sample, data on
some activities of firms are not as good as expected for example there are many missing value
concerning firm's R&D activity partly because Vietnamese firms have not really seen the role of
innovation on firm's value development thus spending on this activity. In general, the study uses the
following data such as:
- Firm types cover state-owned and non-state-owned firms
- Firm age
- Labor cost: The labor cost is composed of the two main items. For the first item, compensation of
employees covers costs relating to wages, salaries, bonus, gratuities and the like considered as salaries;
social security contributed to employees; other compensation out of productions costs. The second item
is the contributions to insurance and pension, health, trade union.
- Total asset at the end of year
- Fixed asset at the end of year
- Depreciation at the end of year: This indicator is identified by the sum of three types of accumulated
depreciation of three long term assets such as tangible fixed assets, financial rented out fixed assets and
intangible fixed assets
- Owner capital at the end of year
- Total revenue of firm: Firm's total revenue is made up of three main items such as turnover of goods,
services activities, turnover of financial activities and other activities.
- Number of employees at the end of year
- Net fixed assets
- Firm's labor productivity
6.4.3.2. Data treatment
We add observations from rice and wood and wood products sectors to those of the footwear sector. We
take the sum of the value of each export contract to obtain the total value of exports of a firm in a given
year. We convert the value of export contracts into VND at the official exchange rate valid at the time of
custom declaration.
We merge data on firms' export value with the information on firms' performance and characteristics. In
2008, for the three sectors, there were 1534 firms among which wood and wood products firms
dominated with 1234 firms, followed by 234 footwear firms and rice firms accounted for fewest with 83
firms active in the same year. (see Appendix 3.1, 3.2, 3.3).
111 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
6.5. Empirical results
6.5.1. OLS regression results
We started analyzing the effect of determinants on firms' export intensity by a model regression with
pooled data then attempting to carry out the regression on each separate sector in order to evaluate
whether those determinants affect firms' export performance differently. As explained, we regress
independent variables with the logit transformed response variable using the OLS method. Then by
generating predictions for our regression model and applying the invlogit function to express the
predictions in units of response variable, we detected that the predictions of the three separate models
were properly constrained to the unit interval though the means are not exactly the same (see Appendix
9.1, 9.2, 9.3). In general, the signs of all coefficients across models are consistent, their magnitudes, in
some cases, show considerable difference. It should be noted that all standard errors of coefficients have
been adjusted for robustness resolving for the problem of heteroscedasticity.
Firstly, the wage cost of firms negatively affects a firm's export intensity in the footwear and wood and
wood products sectors with a 1% level of significance, implying that highly skilled workers are not
actually needed in these low-technology firms. However, this indicator positively but insignificantly
impacts on rice firms' export intensity. Theoretical and empirical studies clearly show that effect of a
worker's wage on trade depends on the country of origin of exports and the kind of products exported.
To be more specific, a high wage per worker is more likely to encourage the trade of developed
countries whereas it reduces the trade of developing and less-developed countries. Wagner (1995)
proved that the export/sale ratio significantly increases with human capital intensity in almost all
industrial sectors of German firms. For British manufacturing firms, the average labor remuneration is
also found to be positively correlated with the probability of a firm being an exporter (Greenway et al.
2004). One possible explanation for the positive effect of firms' wage in Britain is that highly skilled
workers improve firm's competitive position in international markets. However, the negative sign of
firms' wage cost is found in almost all studies examining its effect on the export performance of
developing countries such as studies by Papadogonas et al. (2007), Michiel (2002), Kuma (1994). A
possible explanation is that low-cost labor remains a competitive advantage for these countries.
Literature on exports shows the advantage to a firm's export performance brought about by the
economies of scale in that the larger the size of firm, the lower the unit input cost thus creating a
consistently positive correlation with export performance. Our result is in line with many research
studies such as Aggrey et al. (2010), Sterlacchini (2001) and Papadogonas et al. (2007) showing that
export intensity increases with firm size for the footwear and wood and wood products sectors while
decreasing insignificantly with rice firms. However, for wood and wood products firms, the positive
effect of firm size decreases with size.
Advanced technology is also an important factor to improve productivity of manufacturing firms.
Theoretically, capital intensity also reflects the level of firm's technological investment. Therefore, an
increase in the net fixed assets induces the growth of export value (Özcelik and Taymaz, 2003; Rankin
et al. 2005; Wakeline, 1997; Michiel, 2002).
112 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Nevertheless, it has been found that for Vietnam, capital intensity has consistently reduced the
propensity to export of footwear firms implying that these low-tech and labor-intensive products
do not require much capital for upgrading machinery. Notably, Wagner (1995) also finds that
capital intensity negatively influenced export intensity of firms active in the basic goods industry.
The firm's wage continues to confirm that the degree of modernization of capital stock in basic
good industries is very low.
Table 17. Determinants of export/revenue: OLS regression results (Dependent variable: logit-export intensity, robust s.e)
Independent variables Pooled
data
Footwear Rice Wood and wood products
(1) (2) (3) (4)
Lnwage -0.928*** -1.407*** 0.377 -0.895***
(-5.86) (-3.26) (1.33) (-4.80)
Lnemployee (firm size) 0.171*** -1.480*** -0.481 0.777**
(2.60) (-3.12) (-1.62) (2.19)
Lnemployee2
0.001
(0.01)
0.187***
(4.49)
-0.081
(-0.55)
-0.0692*
(-1.76)
Capintensity -0.002*** -0.0017*** 0.00563 -0.00074
(-6.14) (-6.16) (0.98) (-1.59)
Stateowner 0.0194 0.925 -0.462 -0.176
(0.11) (1.54) (-0.71) (-0.93)
Lnfirmage -0.550*** -0.490 0.0526 -0.586***
(-4.53) (-1.33) (0.14) (-4.30)
Central . . . .
. . . .
North -1.949*** 0.766 -0.774 -2.167***
(-6.43) (0.52) (-0.85) (-6.52)
South -0.581** 1.347 -0.123 -0.604**
(-2.50) (0.98) (-0.16) (-2.51)
Constant 1.812*** 3.081 -0.862 0.868
(3.01) (1.31) (-0.54) (0.85)
Observations 1554 237 83 1234
R2 0.11 0.21 0.11 0.10
F test (p-value)
Ramsey test (p-value)
27.5***
0.001
7.53***
0.74
2.01*
0.81
2.59***
0.007
Notes: * significant at 10%; ** significant at 5%; *** significant at 1%. Absolute of t statistics are in parentheses
Type of firm ownership is also of interest in this chapter since the role of state-owned enterprises
(SOEs) is of crucial importance for the continuing development of Vietnam's trade and for trade policy
113 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
recommendations. In fact, SOEs in Vietnam are an official priority objective therefore benefit from
financial grants from the government budget, lower corporation tax rate, and easy access to state funds
and real estate.
Although, the empirical regression results shows a positive sign of "stateowner" dummy for the
footwear sector and a negative sign for the remaining sectors, these signs are not in fact
significant. However, according to studies on firms' exports, foreign firms have been found to be
more effective in enhancing their export performance (Aggrey et al. 2010; Özçelic and Taymaz,
2003; Farole and Winkler, 2011; Michiel, 2002).
A significantly negative sign for firm age is only found for wood and wood products sector. In fact,
many studies on determinants of exports show no effect of firm age with respect to export performance
(Sousa and Bradley, 2009; Papadogonas et al 2007; Rankin et al. 2006; Robson et al. 2012; Iyer, 2010).
There are few papers finding a positive effect of firm age on export intensity such as Javalgi et al.
(2000), Jongwanich and Kolpaiboon (2008), Amornkitvikai et al. (2012). In other cases, a negative sign
is found for firm age (Tapia et al. 2010; Farole and Winkler, 2011). According to Caparas's paper
(2006), there is a positive linear and negative non-linear relationship between firm age and export
performance for the Philippine clothing and electronics sectors. There are some possible explanations
for the mixed effect of firm age. In fact, firm age reflects a firm's experience in that long-established
firms are more likely to accumulate managerial skills, financial capacity and understanding of the law of
foreign markets. Likewise, young firms may not have enough experience to compete with their larger
international rivals when they launch a global competition campaign. Although, older firms should be
more efficient through their learning-by-doing process (Amornkitvikai et al. 2012), younger firms tend
to be more dynamic, thus finding it easier to adapt to changes in the law and business environment
overseas.
Regression results in the Table 17 clearly show that firms of the wood and wood products sector located
in the northern and southern regions of Vietnam have a significantly decreased propensity to export
compared with firms active in the central region. However, for footwear and rice firms, we do not
have sufficient evidence to conclude the effect of the location on their export performance.
6.5.2. Quantile regression results
6.5.2.1. Quantile regressions for footwear firms
We compared different quantile models by regressing the transformed export intensity with identified
independent variables. As earlier noted, the original response variable values are mainly concentrated
within the range (0, 0.2] and its covariates' effects vary significantly across quantiles (see Appendix
10.1, 10.2, 10.3). We applied quantile regression for quantiles smaller and larger than 50th whereby
can be identified at 10th, 25th, 50th, 75th and 90th. Table 18 reports and compares regression results of
OLS and five different quantile models for footwear firms. In general, except for a firm's capital
intensity, the effects of the remaining covariates change considerably across quantiles and the
114 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
magnitude of the effect at median differs from the effect of mean. We also bootstrapped the standard
errors because we were concerned about heterogeneity between footwear firms.
Regression results show that firm wage level negatively influences footwear firms' export intensity thus
an increase in the wages of those firms does not mean that their propensity to export increases. At all
quantiles the effect of firms' wage is significant and the median estimate is quite far from the OLS
effect. It is important to note that the magnitudes of firms' wage effect at 25th and 90th are similar and
there is a fluctuation in the magnitudes of firms' wage effect from lower to upper quantiles.
Consistent with the OLS regression for firm size, firm size estimate effects at different quantiles are
positive to export intensity implying that footwear firms can increase exports by expanding their
economies of scale. The magnitude of firm size effect at median is more or less similar to the OLS
effect. From 25th to 90th quantile, the magnitudes of firm size do not significantly change, nonetheless
at 10th quantile, it is fairly small compared with the others and is not significant.
It should be noted that a positive and significant effect of state owned firm variable is found at 25th
quantile indicating that a prioritized entity such as state owned firms of footwear sector significantly
contributes to their propensity to exports. Meanwhile, older footwear firms affect their export
performance negatively but the effect is not significant across quantiles.
115 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Table 18. Models of export intensity via OLS and quantile regression (footwear firms)
(1) (2) (3) (4) (5) (6)
Independent variables OLS Q(0.10) Q(0.25) Q(0.50) Q(0.75) Q(0.90)
Lnwage -1.479*** -1.429*** -0.956* -1.722*** -1.122* -0.989*
(-3.56) (-2.60) (-1.94) (-3.95) (-1.81) (-1.71)
Lnemployee (firm size) 0.539*** 0.116 0.584** 0.693*** 0.654*** 0.515*
(3.21) (0.33) (2.49) (2.93) (3.17) (1.91)
Capintensity -0.00018*** -0.00012** -0.00021 -0.00014 -0.00021 -0.00025
(-6.10) (-2.45) (-0.16) (-1.20) (-0.79) (-0.73)
Stateowner 0.745 1.137 1.228* 0.525 0.473 -0.204
(1.31) (1.01) (1.76) (0.79) (0.63) (-0.26)
Lnfirmage -0.521 -0.951 -0.899 -0.219 -0.091 -0.129
(-1.43) (-1.44) (-1.52) (-0.50) (-0.22) (-0.24)
Constant -0.143 -2.035 -4.148** -0.751 0.146 2.617
(-0.08) (-0.68) (-2.09) (-0.29) (0.06) (0.79)
Observations 237
R2 0.204
Pseudo R2 0.15 0.10 0.12 0.10 0.09
Notes: * significant at 10%; ** significant at 5%; *** significant at 1%. Absolute of t-statistics of OLS regression are in parentheses.
Coefficients and standard errors of quantile regressions are bootstrapped. Bootstrap is performed with 20 replications.
116 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
6.5.2.2. Quantiles regression for rice firms
The number of rice firms was least compared with the other two firms sectors, only 83 such firms in
2008. In general, only the magnitude of a firm's capital intensity at median is similar to OLS estimate
whereas those of the remaining coefficients at median are quite different from the OLS estimates.
Moreover, almost all effects of the model coefficients are not significant with respect to the OLS as well
as to the quantile regressions.
In contrast with the negative effect of footwear firms' wage, it is found to be positive for the OLS
regression and across quantiles except for the 10th quantile. Surprisingly, there is a significant and
positive impact of firm size on rice firms' export intensity at the 90th quantile. Moreover, the size of rice
firms decreases at all quantiles but this effect is only significant at 10th quantile suggesting that large
firms discourage their export intensity.
We found a positive sign for the capital intensity effect on the OLS and lower quantiles estimates,
whereas at the upper quantiles, an intensive investment in firms' fixed assets has a negative impact on
their export performance. However, this effect is found not to be significant at all quantiles. The effect
of the state-owned firm variable is negative for OLS and quantile regression, nonetheless it is found to
be insignificant. In particular, the magnitudes of this coefficient are similar at the lower and upper
quantiles at 0.3 and 0.4 respectively.
A long presence in the market does not necessarily mean firms can work effectively in foreign markets.
The firm age variable gives a positive and negative signs at the lower and upper quantiles respectively
but its coefficients are not significant except for the median estimate. Generally, firm age is not an active
factor stimulating their propensity to export compared with the younger ones.
117 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Table 19. Models of export intensity via OLS and quantile regression (rice firms)
(1) (2) (3) (4) (5) (6)
Independent variables OLS Q(0.10) Q(0.25) Q(0.50) Q(0.75) Q(0.90)
Lnwage 0.437 -0.350 0.466 0.778 0.563 0.496**
(1.49) (-0.42) (0.70) (1.40) (1.21) (2.33)
Lnemployee (firm size) -0.493* -0.984** -0.359 -0.147 -0.128 -0.115
(-1.69) (-2.12) (-1.04) (-0.47) (-0.35) (-0.47)
Capintensity 0.00057 0.0022 0.0012 0.00072 -0.00043 -0.00116
(1.04) (1.56) (1.08) (0.69) (-0.45) (-0.47)
Stateowner -0.368 -0.335 -0.394 -0.251 -0.462 -0.410
(-0.58) (-0.17) (-0.43) (-0.26) (-0.48) (-0.60)
Lnfirmage 0.0130 0.467 0.172 -0.707 -0.236 -0.299
(0.04) (0.84) (0.42) (-1.12) (-0.52) (-0.68)
Constant -1.513 -0.852 -4.008 -2.692 -1.221 0.172
(-0.93) (-0.31) (-1.37) (-1.26) (-0.84) (0.13)
Observations 83
R2 0.10
Pseudo R2 0.13 0.10 0.06 0.04 0.08
Notes: * significant at 10%; ** significant at 5%; *** significant at 1%. Absolute of t-statistics of OLS regression are in parentheses.
Coefficients and standard errors of quantiles regression are bootstrapped. Bootstrap is performed with 20 replications.
118 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
6.5.2.3. Quantiles regression for wood and wood products firms
The effect of wood and wood products firms' wage cost is similar to footwear firms in that an increase in
this cost does not encourage firms' propensity to export. As clearly shown in Table 20, the magnitudes
of firms' wage are very different between the OLS estimate and the median estimate. At the lower
quantiles, the magnitudes of wage cost are also smaller than those at the upper quantiles, nonetheless
they are more significant. At all quantiles, significant effect of firms' wage cost is found. A negative
effect of wood and wood products firm wage cost across quantiles continues to demonstrate Vietnam's
competitive advantage through its lower cost of labor.
In line with footwear firms, firm size in the wood and wood products sector positively influences export
intensity though the magnitudes are apparently smaller. However, at 10th quantile, firm size effect is
negative and higher levels of quantiles when higher weights are given to export intensity, firm size
effect is the weakness compared with it at median level. A highly positive effect of firm size again
verifies the theory of economies of scale, which play an important role for firms enabling them to
compete in international market.
Although the capital intensity is negative in almost all regression models, it is only significant at 50th and
75th quantile. At 90 percentile, this indicator is found to be positive but insignificant. This finding shows
that Vietnam's wood and wood products exporting firms remain a labor-intensive sector, as shown in
some other studies for the negative effect of capital intensity, such as Papadogonas et al. (2007), Wagner
(1995) and Bleaney and Wakelin (2007).
In contrast with the effect of the SOEs in the footwear sector, in the wood and wood products sector,
SOEs are found to be less efficient compared with non-state owned firms. Although, the effects of SOEs
are negative on firms' export intensity in almost all regressions, they are not significant across quantiles.
This finding strongly confirms the unimportant role of such a firm in encouraging their propensity to
export.
The firm age variable shows a negative and significant effect on export intensity for all regressions
except for 0.10th quantile which is found to be insignificant. The finding is in line with papers of Tappia
et al. (2010), Farole and Winkler (2011). The negative effect of firm age confirms that older firms of
three sector considered are not relevant for the international trade.
119 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Table 20. Models of export intensity via OLS and quantile regression (wood and wood products firms)
(1) (2) (3) (4) (5) (6)
Independent variables OLS Q(0.10) Q(0.25) Q(0.50) Q(0.75) Q(0.90)
Lnwage -0.961*** -1.566*** -1.355*** -0.694*** -0.316** -0.248**
(-5.37) (-5.23) (-7.16) (-4.28) (-2.18) (-2.02)
Lnemployee (firm size) 0.208*** -0.338** 0.269 0.600*** 0.379*** 0.184***
(2.74) (-2.11) (1.44) (8.06) (7.53) (2.88)
Capintensity -0.00077 -0.00091 -0.00070 -0.0012** -0.0012* 0.000021
(-1.56) (-0.53) (-0.58) (-2.55) (-1.87) (0.03)
Stateowner -0.129 0.263 -0.0597 -0.222 -0.363 -0.147
(-0.68) (0.66) (-0.18) (-0.85) (-1.64) (-0.88)
Lnfirmage -0.667*** -0.459 -0.729*** -0.719*** -0.397** -0.412**
(-4.81) (-1.43) (-3.70) (-5.82) (-2.44) (-2.32)
Constant 1.458** 1.140 0.594 -0.495 0.676 2.371***
(2.21) (1.11) (0.72) (-0.73) (1.17) (3.89)
Observations 1234
R2 0.075
Pseudo R2 0.07 0.06 0.06 0.03 0.01
Notes: * significant at 10%; ** significant at 5%; *** significant at 1%. Absolute of t-statistics of OLS regression are in parentheses.
Coefficients and standard errors of quantiles regression are bootstrapped. Bootstrap is performed with 20 replications.
120 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
6.6. Conclusion
This chapter analyzes the effects of firms' performance and characteristics on their export intensity in the
labor-intensive manufacturing sectors, footwear, rice and wood and wood products considered to be at
an advantage in the international market.
We are particularly interested in identifying determinants on export activities of both manufacturing and
non-manufacturing firms of three selected sectors involved in export activities for 2008. Although the
OLS approach enables us to examine effects of firms' characteristics on their export intensity, which is
suitable for logit transformation, the quantiles regression method is also used because it yields richer
characterization of the data. In this respect, we observe that firms are highly heterogeneous within a
single sector and there are numerous extreme observations of firms' export intensity lying in the (0, 0.2]
interval.
The regression results of OLS and quantiles show that almost all the mean and median estimates of
independent variables are very different except for firms' capital intensity, thus confirming our concern
about the heterogeneity of firms. Another point to note is that a significant sign in rice firms'
determinants is infrequent for both OLS and quantile regressions.
The export intensity is negatively affected by the wage, age and capital intensity of these firms. These
findings confirm the fact that Vietnam's exports are still concentrating on low labor cost, labor intensive
and low-tech production. Firms' sizes continue to confirm their role in increasing their propensity to
export.
These findings suppose that policy measures need to promote extension of economies of scale for
footwear and wood and wood products firms through tax incentives whereas the economies of scale of
rice firms do not required further promotion.
Although low cost of labor may be regarded as a competitive advantage of many manufacturing sectors
of Vietnam including footwear and wood and wood products manufacturing industries, its contribution
to the global value chain is considerable. This is partly because the majority of export activities are in
the form of processing contracts whereby Vietnamese firms mainly import machinery or spend at low-
tech level for product processing. In order for Vietnam to achieve a higher position in the global value
chain and compete better in the international market, policy measures need to promote high-tech
industries, which in turn encourage more highly skilled and better-paid workers.
Export literature shows that highly productive firms are more likely to be involved in the international
market. This concept is also true for a developing country such as Vietnam, where firms' export intensity
accounts for only a small percentage of their total revenue. Moreover, the wage cost magnitudes of wood
and wood products firms decrease sharply at the lower quantiles compared with the upper quantiles
showing that where export intensity is highly concentrated, a high percentage of low cost labor is used.
In order to promote exports, this means that government officials should consider measures to improve
labor skills to adapt them to high-tech production needs.
121 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
CHAPTER 7. CONCLUSION
SUMMARY
This chapter summarizes the main findings of the dissertation and suggests some export trade
implications for policy makers. The scope of research and its limitation are accordingly addressed.
Further research and exploration indeed should be carried out in order to obtain better guidance for
exporting entrepreneurs in remaining sectors.
122 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
7.1. Major findings and lessons retrieved through the empirical analysis
- Vietnamese export structure has been dramatically transformed and demonstrates the spirit of a deeper
integration of Vietnam into the world economy as well as the advantages of pursuing a trade liberalization
policy.
Every country finds the process of globalization and economic integration to be an important
opportunity to develop its production and trade network far beyond its own borders. For a small and
developing country such as Vietnam, especially which had erroneously pursued a centralized state
planning model, trade liberalization and economic integration eventually became an inevitable option.
In almost 30 years following Doi Moi, Vietnam has made outstanding progress in trade with a very
impressive gain in export revenue of 114.5 billion USD (at 2012 value). Especially, the number of
Vietnam' trading partners reached 220 countries in 2009 starting from 33 partner countries in 1986
marking a large step towards integration. Moreover, trade liberalization has also encouraged through
multilateral and bilateral trade agreements between Vietnam and other major economic associations
ASEAN, EU and WTO as well as with a single country for example Vietnam-USA trade agreement.
Chapter 3 has depicted the structure of exports and country partners which continues to change since the
country's reform. In the early years of economic reform, the contribution of western partner countries
such as the EU and USA was very limited only accounting for nearly 5% and 0% respectively. However, in
recent years, especially since the signing of the Vietnam-US trade agreement in 2000, both trading
partners became extremely important with an export share of 17.1% and 16.8% respectively in the
2002-2012 period. Moreover, there has been an impressive increase in the number of export product
categories with the largest increase being in manufactured products compared with primary ones. APEC
has emerged as the most significant of the major importing country groups for both primary and
manufactured products.
In fact, as Vietnam's export revenue from manufactured product increased, this has had a positive effect
on the country's export structure. Vietnam's exports traditionally covered relatively labor-intensive products
such as clothing and textiles, footwear, embroidery as well as agricultural commodity such as rice, coffee,
pepper and cashew nuts. For the nearly 30 years since the country began its economic reforms, Vietnam
has retained its comparative advantage because of its lower labor costs.
- Although Vietnam has many importing partners for the 1997-2009 period, the 10 largest partners out
of over 200 countries account for 69% of Vietnam's total export revenue.
- Landlocked countries continue to be generally regarded as less favorable markets for exporting
countries and particularly in the case of Vietnam where most exports rely on maritime transport, which
require a seaport at final destination the absence of which will inevitably give rise to additional cost.
- Macro level export data and micro level data may give contradictory findings hence firm level exports
may not necessarily be affected in the same way as the country's exports.
123 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Generally, the findings of chapter 4 show that a country economic size determines its attractiveness to
exporting countries. However, not all exporting firms take this for granted particularly firms active in
the rice and wood and wood products sectors. In practice, this means such firms based their export
decision on anticipated net profit on a particular transaction.
Wealthier countries constitute a stimulus factor for Vietnamese footwear exports. However, for the rice
and wood and wood products exports, the signs are also positive but insignificant thus indicating
insufficient evidence to confirm that these firms consider richer countries as attractive.
In general, a country finds that geographical distance is an obstacle to trade. However, it is not
necessarily a problem for firms in certain sector such as footwear and wood and wood products, which
shows that transportation costs do not always impede exports. In fact, geographical distance factor gives
a negative and insignificant sign for Vietnam’s total exports, it turns out to affect positively and
significantly exports of the footwear and wood and wood products firms but affect negatively the rice
firms’ exports
Vietnam's membership of ASEAN ensures preferential treatment for tax and import license procedures
at both country and firm level except for wood and wood products firms, the export volume of which to
the ASEAN countries is relatively smaller than to non-ASEAN countries.
- Chapter 5 also tests effects of gravity factors which are included in chapter 4 on firms’ exports in
footwear, rice and wood and wood products sector. Almost all firms in the three export sectors
considered use FOB contracts for exports where it is the importer's responsibility to cover charter and
insurance costs and Vietnamese exporting firms only charge the importers the product cost, domestic
transportation costs and any other costs, which may arise within Vietnam. If Vietnamese firms could
absorb additional costs arising outside Vietnam's border, they would be better able to compete on
international markets.
- Chapter 6 analyzes the role of firm's location obviously varies between export sectors because of the
intrinsic characteristics of its production.
While footwear and rice firms are mainly concentrated in the northeast and southwest of Vietnam, wood
and wood products firms tend to be spread throughout the central and highland regions since these firms
need to locate near the production site of their raw material. Therefore, export revenue of wood and
wood products firms located in these regions accounts for almost all the total export revenue of this
sector. Nevertheless, the density of export value by the footwear and rice sectors does not entirely
correspond to their region in which they are mainly concentrated. In fact, wood and wood products firms
do not necessarily find the north of Vietnam a stimulus factor when considering an increase in the share
of exports in their total revenue. In general, firm locations are significant for wood and wood products
sector but insignificant for the other two sectors.
- Firm age is generally irrelevant for the export intensity of wood and wood products sector.
124 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
In certain cases, more experienced older firms grow larger therefore are better able to benefit from
economies of scale. In other cases, younger firms being smaller can more easily keep up with the
volatile and dynamic international market. Our research shows that older firms are not always eager to
improve their export performance.
- In general, higher wages do not increase the propensity to exports. However, in the case of Vietnam,
higher wages have reduced the propensity to exports for footwear and wood and wood products sectors
indicating that low labor costs retain their role in Vietnam's competiveness.
- Studying the role of state-owned enterprises is very important in a small emerging country such as
Vietnam, which has experienced a long period of relative economic decline under a state centrally
planned system. However, we find that state-owned exporting firms underperform compared with non-
state firms for rice and wood and wood products sectors. Even though such firms may have received
preferential treatment from the government, they have been found to be less efficient and have poorer
export performance.
7.2. To what degree does the dissertation answer the research questions and what are its
limitations?
The dissertation has described the overview of Vietnamese exports and explained systematically the
country's export trade pattern at country and firm's level by applying standard gravity factors. The
dissertation accordingly has attempted to answer questions on the determinants of Vietnam's exporting
firms in the world market.
The presentation of the theoretical background of international trade is of the utmost importance and
attempts to explain why and how countries and firms trade. The international trade paradigm also
focuses on the role of the country and firm in the international market as these two entities have their
own reasons to trade. As such, the dissertation attempts to show how Vietnam operates in international
market compared with its firms.
Vietnam's economic reform process is highly challenging in two respects. Firstly, the country's trade
policy centers on liberalization and the government is attempting to further develop a legal framework
for firms intending to penetrate and to gain a larger market share in foreign markets. Chapter 4 and 5
have answered the question of whether economic integration into the ASEAN free trade area could have
positive effect on the export trade of Vietnam as well as exports of firms in the three mentioned sectors.
Vietnam's membership of ASEAN is beneficial to its total export revenue and consequently to the
country as a whole. Depending on individual characteristics, various export firms benefit to different
degrees. Nonetheless, firms active in the wood and wood products sector do not find ASEAN an
attractive factor.
Secondly, as part of this process, firms examine which internal factors would affect their export
performance then devise further measures to cope with tougher competition in the international market.
Although, firms face constant challenges everywhere, nevertheless exports continues to grow. Chapter 6
125 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
has successfully helped resolve frequent challenges of firms such as potential economies of scale,
capital and human intensity.
We have also attempted to address the question of chapter 4 and 5 is whether geographical distance (a
proxy of transport cost) matter. In fact, chapter 4 shows a negative effect of distance on Vietnam’s total
exports whereas chapter 5 indicates that it does not matter for exports of footwear and wood and wood
products firms. Especially, a positive effect of contiguous countries on Vietnam’s exports is found in
chapter 4 whereas individual firms such as rice do not consider contiguous countries supportive
However, the dissertation has the following limitations:
- Although our dissertation describes the whole of Vietnam's trade activity for the long period since
implementation of reforms in certain areas namely changes in export commodities structure, the
diversity destination markets and the national comparative advantage as expressed by its Herfidale index
but it does not explain the causes of such changes due to data limitations.
- The most interesting feature of our dissertation is that it attempts to examine the effects on firms'
export performance of gravity factors considered as external factors. However, these factors do not have
the same effect for all firms sectors and therefore no general conclusion can be drawn. Similarly, this
limitation applies to the effects of firm's internal factors because the dataset only covers firms in three
sectors mentioned.
- If our dissertation was to be able to analyze the firm size factor based on certain categories, possible
different outcomes of this determinant may lead us to more specific conclusions on firms' export
performance.
- Although, the Vietnamese enterprise survey included questions about firms' expenditure on research
and development38
, the very low response level meant that insufficient data were available to be able to
draw conclusion on these activities.
- Although our dissertation found that geographical distance did not necessarily undermine a firm's
export performance, it could not identify the precise factor determining such circumstances.
- Another important limitation is that we do not have the relevant data for such firms’ factors as R&D.
7.3. Implications for Vietnam's export trade
Our dissertation mainly focuses on certain external factors of Vietnamese exports at macro and micro
level and analyzes some important determinants of exporting firms in three sectors namely footwear,
rice and wood and wood products. Relevant findings based on empirical analysis serve to suggest policy
implications and firms' managerial practices.
- Our regression results on Vietnamese trade patterns continue to confirm the stimulus effect of
ASEAN on export performance since tariff exemptions, tax cuts or other non-tariff trade barriers are
38
Costs relate to firms' research and technology innovation activity
126 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
eliminated by its member countries. It is planned that the ASEAN Economic Community (AEC)39
will
be established late 2015 with the ultimate goal of full economic integration of the community. Major
methods that ASEAN needs to apply to construct a single ASEAN market include common product
quality standard, rules and mechanisms on custom clearance and common rules of origin. It is important
to note that under AEC rules, ASEAN members have committed to removing taxes 3 years earlier than
their previous commitment arising from the ASEAN Free Trade Area (CEPT/AFTA). As such, this
enables the Government to take measures to encourage firms to prepare for further integration
particularly in priority export sectors such as agriculture, seafood, wood and wood products, garments,
electronics. Possible methods include providing training courses for firms so that they can better
recognize opportunity and threat from AEC membership, issuing legal documents, which comply with
the AEC rules for firms, and creating a more convenient legal framework for both State and non-state
entrepreneurs.
However, the role of ASEAN is not confirmed for wood and wood products sector even exports of these
products are depressed if they are supplied to this area. Managers of wood and wood products firms
need to note that ASEAN importing market should not be focused.
- In general, wealthier nations are preferred destination market for Vietnamese exports because of their
higher purchasing power. The government of Vietnam therefore tries to negotiate proactively and
conclude trade agreement as well as memorandum of understanding with such countries as these
measures facilitate trade for local firms.
In particular, managers of footwear firms can decide to choose high income destination markets to
export as well as propose promotion method for these products provided to these markets However,
management boards of rice firms do not consider high income country to be their destination market. In
fact, there are several reasons for firms to enter international market and to sign contracts such as to
improve their net profits, to expand their market share or to develop potential markets.
- Despite the generally negative effect of geographical distance, footwear and wood and wood products
firms are successful in distant markets particularly where such markets have high purchasing power. As
such, managers of those firms could consider distant importing countries as the target markets of
increasing their exports.
- One important empirical result for a firm's determinants is that higher wages reduce a firms' propensity
to export in general which significantly affects footwear and wood and wood products firms. Although,
in some cases, increasing wages somewhat improve worker's productivity thus increasing export
revenue, in other cases for economic reasons, higher wages can not be afforded by employers for
manual workers in manufacturing sectors where skilled labor is not essential.
- The size of firms affects differently their export performance of the three sectors. As a matter of policy
implications, for the rice firms sector, the finding indicates that small firms should be encouraged to
39
Along with AEC, ASEAN covers the other two important columns such as ASEAN Security Community (ASC)
and ASEAN Culture-Society Community (ACSC)
127 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
increase their export intensity because there is a negative association between the size of rice firms and
their export performance. Notably, for the wood and wood products firms, firm size decreases with size
suggesting that large firms do not facilitate their export performance. However, managers of footwear
firms could benefit from the economies of scale so that increasing their firm size is a good decision to
increase exports.
- From the viewpoint of footwear, rice and wood and wood products firms, although these firms are
restricting their fixed assets investment, managerial practices should encourage new fixed assets investment
in new technology and innovation which help improve firms’ productivity. More importantly, in the period
of deep economic integration, Vietnamese firms are inevitably facing with a fierce competition from the
other less developed and developing countries in serving low labor cost products.
- Although, state-owned firms have been granted preferential or priority treatment by the Government,
they appears to be less efficient than others. The regression result shows that state-owned firms in rice
and wood and wood sectors are lagging in export performance. Ongoing Government support for state-
owned firms may lead to an unfair competitive environment especially for small, privately owned firms
since the protection of the Government seem to be unreasonable.
7.4. Some proposals for future further research on export trade
The dissertation, while showing some weaknesses, sought and analyzed the pattern of Vietnamese
exports by replicating the gravity model and investigated determinants for firms' exports. In chapter 4,
the dissertation only used the one-way gravity model, which allows us to examine the effects on
Vietnamese exports of the economic indicators of importing countries. Rather, the same controllable
economic indicators of exporting countries should be incorporated into the regression model so as to
confirm the effects of Vietnam's economic size on its exports. The 1997-2009 period may not be long
enough as a time series for the adequate consideration of the structural changes in Vietnam's export
trade to over 150 countries (see chapter 4). In fact, a qualitative analysis of Vietnamese exports (see
chapter 3) from 1986, the year of economic reform shows a major increase in export revenue as well as
the country's early steps towards economic integration and trade liberalization. Another weakness also
revealed through the fixed effects regression in chapter 4 is that it removes time-invariant variables and
GDP and GDP per labor force may be endogenous because they may be significantly correlated with
the unobserved individual-level random effect. In fact, data has been available on such other indicators
as the rule of law, the control of corruption, the government effectiveness, the political stability and the
voice and accountability index of importing countries as suggested by Egger (2002). However, because
the time series of these indicators which cover only 8 years of the 2002-2009 period meanwhile the time
series of the original model include 13 years, testing these indicators as instrumental variables is not
possible.
Studying the role of individual firms to determine a country's trade pattern is a more effective approach
because it is the more dynamic firms, which self-select for the international markets. Since aggregated
export trade data for a sector or at country level may result in biased estimates, more persuasive
128 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
regression results are obtained from export data at firm level. By studying firms' export trade pattern, the
dissertation enabled discrimination between heterogeneous firms in sectoral basis. However, research
result for exporting firms active in only the three sectors previously mentioned can not be extrapolated
to the trade pattern of all sectors. For future research, once a full dataset on firms is available, more
comprehensive conclusion can be drawn.
Studying the structure of market competition of Vietnamese exporting firms in order to assess how one
firm manages to serve more markets or even fewer markets with larger proportion of export sale than
the other firms is also our interest. In addition to this, the local government of Vietnam may affect these
firms’ competitiveness within and/or outside the country. These future researches should be
implemented when data on firms’ export unit value and the province competitive index reflecting the
government’s efficiency are available.
The effect of geographical distance on country level data is mostly negative but in a few studies as well
as in our dissertation where the gravity model is replicated, the distance is found to have a positive sign
thus contradicting the traditional viewpoint that the higher costs of its exporting to distant countries are
an insurmountable obstacle to trade. As such, future research should dig out what factors determine how
firms overcome additional transport and other costs outside Vietnam. Possible determinants may include
a firm's productivity and its export strategy, which may also explain the success of some firms in distant
markets.
As mentioned above, once full data on firms' exports of in all sectors have been extracted, the author of
the dissertation will attempt to draw conclusions on the determinants for all sectors and compare
coefficients accordingly. We consider innovation indicators would be interesting particularly for firms
located in a developing and emerging country such as Vietnam. Therefore, we need an appropriate
survey method to collect information to resolve the missing data problem of the existing database.
129 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
REFERENCES
Adam C and D. Cobham. 2007. Modelling bilateral trade resistance in a gravity model with exchange
rate regimes, Conference Papers 2007, Centre for Dynamic Macroeconomic Analysis, School of
Economics & Finance, University of ST Andrews of UK.
Aggrey N., L. Eliab and S. Joseph. 2010. Determinants of Exports Participation in East African
Manufacturing Firms, Current Research Journal of Economic Theory, 2(2), 55-61.
Aitken D.N. 1973. The Effect of the EEC and EFTA on European Trade: A Temporal Cross-Section
Analysis, American Economic Review, 63(5), 881-92.
Amornkitvikai Y., C. Harvie and T. Charroenrat. 2012. Factors Affecting the Export Participation
and Performance of Thai Manufacturing Small and Medium Sized Enterprises (SMEs),
57th International Council for Small Business World Conference, 1-36, Wellington, New-
Zealand.
Anderson J.E and E.V. Wincoop. 2003. Gravity with Gravitas, The American Economic Review,
93(1), 170-92.
Anderson J.E 1979. A Theoretical Foundation for the Gravity Equation, The American Economic
Review, 69(1), 106-16.
Arkolakis C and A.L. Muendler. 2009. The Extensive Margin of Exporting Goods: A Firm-level
Analysis, February 2009, University of California, San Diego, unpublished manuscript (URL
econ.ucsd.edu/muendler).
Baltagi H.B., P. Egger and M. Pfaffermayr. 2003. A Generalized Design for Bilateral Trade Flow
Model, Economic Letters, 80(3), 391-97.
Baltagi H.B. 2005. Econometric Analysis of Panel Data, Third edition, John Wiley & Sons, Ltd.
Banga R. 2006. The Export-Diversifying Impact of Japanese and US Foreign Direct Investments in the
Indian Manufacturing Sector, Journal of International Business Studies, 37(4), 558-68.
Bastos P and J. Silva. 2010. The Quality of a Firm's Exports: Where You Export to Matters, Journal of
International Economics, 82, 99-111.
Baum C. 2013. Quantile Regression, Lecture Notes for Applied Econometrics, Boston College.
Bergstrand H.J. 1985. The Gravity Equation in International Trade: Some Microeconomics
Equation, The Review of Economics and Statistics, 67(3), 474-81.
Bergstrand H.J. 1987. The U.S Trade Deficit: A Perspective from Selected Bilateral Trade Models. New
England Economic Review, Federal Reserve Bank of Boston, 19-31.
Bergstrand H.J. 1989. Generalized Gravity Equation, Monopolistic Competition, and the Factor –
Proportion Theory in International Trade, The Review of Economics and Statistics, 71(11),
143-53.
Bernard A.B and J.B. Jensen. 2004a. Exporting and the Productivity in the USA, Oxford Review of
Economic Policy", 20(3), 343-57.
Bernard A.B and J.B. Jensen. 2004b. Why Some Firms Export, The Review of Economics and
Statistics, 86(2), 561-69.
130 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Bernard A.B and J.B. Jensen. 2007. Firm Structure, Multinationals, and Manufacturing Plant Deaths,
The Review of Economics and Statistics, 89(2), 193-204.
Bernard A.B and J.B. Wagner. 1997. Exports and Success in German Manufacturing,
Weltwirtschaftliches Archiv/Review of World Economics, 137(1), 105-23.
Bernard A.B., J.B. Jensen, S.J. Redding and P.K. Schott. 2007b. Firms in International Trade, The
Journal of Economic Perspectives. 21(3), 105-30.
Bernard A.B., S.J. Redding and P.K. Schott. 2007a. Comparative Advantage and Heterogeneous
Firms, The Review of Economic Studies, 74 (1), 31-66.
Bhavani T and S. Tendulka. 2001. Determinants of Firm-Level Export Performance: A Case Study
of Indian Textile Garments and Apparel Industry, Journal of International Trade and Economic
Development, 10(1), 65-92.
Biesebroeck J.V. 2005. Exporting Raises Productivity in Sub-Saharan African Manufacturing Firms,
Journal of International Economics, 67, 373-91.
Bigsten A., P. Collier, S. Dercon, M. Fafchamps, B. Gauthier, J.W. Gunning, A. Oduro, R.
Oostendorp, C. Pattillo, M. Söderbom, F.Teal and A. Zeufack. 2000. Exports and Firm- Level
Efficiency in African Manufacturing Mimeo. Centre for the Study of African Economies at
Oxford University, Oxford.
Bleaney M and K. Wakelin. 2002. Efficiency, Innovation and Exports, Oxford Bulletin of Economics
and Statistics, 64 (1), 3-15.
Bonaccorsi A. 1992. On the Relationship between Firm Size and Export Intensity, Journal of
International Business Studies, 23(4), 605-35.
Brun J. F., C. Carrere, P. Guillaumont and J. de Melo. 2005. Has Distance Died? Evidence from a
Panel Gravity Model, World Bank Economic Review, 19(1), 99-120.
Campa M.J and F.M. Guillén. 1999. The Internationalization of Exports: Firm- and Location- Specific
Factors in a Middle-Income Country, Management Science, 45(11), 1463-78.
Caparas D. 2006. Determinants of Exporting Performance in the Philippine Manufacturing Sector,
Discussion Paper Series No. 2006-18, Makati City, Philippine.
Carrère C. 2006. Revisiting the Effects of Regional Trade Agreements on Trade Flows with Proper
Specification of the Gravity Model, European Economic Review, 50, 223-47.
Cheng I and H.J. Wall. 2005. Controlling for Heterogeneity in Gravity Models of Trade and
Integration, Federal Reserve Bank of St. Louis Review, 87(1), 49-63.
Christie E. 2002. Potential Trade in Southeast Europe: A Gravity Model Approach. Working Paper,
the Vienna Institute for International Economic Studies-WIIW downloaded on 14 March 2002.
Clerides S., S. Lach and J. Tybout. 1998. Is Learning by Exporting Important? Micro-Dynamic
Evidence From Colombia, Mexico and Morocco, Quarterly Journal of Economics 113, 903-48.
Crozet M and P. Koenig. 2010. Structural Gravity Equations with Intensive and Extensive Margins,
Canadian Journal of Economics, Canadian Economics Association, 43(1), 41-62.
Darvas Z. 2012. Real Effective Exchange Rates for 178 Countries: A New Database, Bruegel working
paper.
Deardorff V. 1995. Determinants of Bilateral Trade: Does Gravity Work in a Neoclassical World?
131 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
NBER Working Paper 5377, National Bureau of Economic Research, Inc.
Delgado W.A., J.C. Fariñas and S. Ruano. 2002. Firm Productivity and Export Markets: A Non-
Parametric Approach, Journal of International Economics, 57, 397-422.
Do T.T. 2006. A gravity model for trade between Vietnam and twenty-three European countries.
Department of Economics and Society, Dalarna University, Sweden.
Doan T.T.H and K. Kiyota. 2014. Firm-Level Evidence on Productivity Differentials and Turnover in
Vietnamese Manufacturing, Japanese Economic Review, 65(2), 193-217.
Egger P and M. Pfaffermayr. 2003. The Proper Panel Econometric Specification of the Gravity Equation
Equation: A Three-Way Model with Bilateral Interaction Effects, Empirical Economics, 28, 571
-80.
Egger P. 2002. An Econometric View on the Estimation of Gravity Models and the Calculation of
Trade Potentials, The World Economy, 25(2), 297-312.
Elliott D.R. 2007. Caribbean regionalism and the Expectation of Increased Trade: Insights from a
Time Series Gravity Model, International Trade Economic Development, 16(1), 117-36.
Fakih A and L. P. Ghazalian. 2013. Why Some Firms Export? An Empirical Analysis for Manufacturing
Firms in the MENA Region. IZA DP No. 7172.
Farole T and D. Winkler. 2011. Firm Location and the Determinants of Exporting in Developing
Countries, International Trade Department of the World Bank, Policy Research Working
Paper 5780.
Feenstra R.C., J.R. Markusen and A.K. Rose. 2001. Using the Gravity Equation to Differentiate
Among Alternative Theories of Trade, Canadian Journal of Economics, 34(2), 430-47.
Fernandes A.M and A.E. Isgut. 2005. Learning-by-Doing, Learning-by-Exporting, and Productivity:
Evidence from Colombia, Policy Research Working Paper Series 3544, the World Bank.
Filippini C and V. Molini. 2003. The Determinants of East Asian Trade Flows: A Gravity Equation
Approach, Journal of Asian Economics, 14(5), 695-711.
Fitzsimons E., V. Hogan and J.P. Neary. 1999. Explaining the Volume of North-South Trade in
Ireland: A Gravity Model Approach, The Economic and Social Review, 30(4), 381-401.
Frankel J.A and A.K. Rose. 2000. An Estimate of the Effect of Currency Unions on Trade and Growth,
Alesina-Barro conference, Stanford University.
Freeman J., C. Styles and M. Lawley. 2012. Does Firm Location Make a Difference to the Export
Performance of SMEs?, International Marketing Review, 29(1), 88-113.
Freeman J., M. Lawley and C. Styles. 2010. Does Firm Location Influence the Export Performance
of Australian SMEs?, 2010 Conference of the Australian and New Zealand Marketing Academy
(ANZMAC).
Geraci V.J and W. Prewo. 1977. Bilateral Trade Flows and Transport Costs, The Review of
Economics and Statistics, 59(1), 67-74.
Glick R and A.K. Rose. 2002. Does a Currency Union Affect Trade? The Time-Series Evidence,
European Economic Review, 46, 1125–51.
Goldstein M and M. Khan. 1978. The Supply and Demand for Exports: A Simultaneous Approach, The
Review of Economics and Statistics, 60(2), 275-86.
132 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Greenway D., N. Sousa and K. Wakelin. 2004. Do Domestic Firms Learn to Export from
Multinationals?, European Journal of Political Economy, 20, 1027-43.
Hausman A.J. 1978. Specification tests in econometrics, Econometrica, 46(6):1251–72.
Heckscher E. 1919. The Effect of Foreign Trade on the Distribution of Income, Ekonomisk Tidskrift, 21,
497–512.
Helpman E and P.R. Krugman. 1985. Market Structure and Foreign Trade. Increasing Returns,
Imperfect Competition, and the International Economy, Cambridge, MA: MIT Press.
Hillberry R and D. Hummels. 2008. Trade Responses to Geographic Frictions: A Decomposition
Using Micro-Data, European Economic Review, 52, 527-50.
Honoré B.E. 1992. Non-linear Models with Panel Data, cemmap working paper CWP13/02, the
Institute of Fiscal Studies Department of Economics, UCL.
Iyer K. 2010. The Determinants of Firm-Level Export Intensity in New Zealand Agricultural and
Forestry, Economic Analysis & Policy, 40(1), 75-86.
Javalgi G.R., D.D. White and O. Lee. 2000. Firm Characteristics Influencing Export Propensity: an
Empirical Investigation by Industry Type. Journal of Business Research, 47(3), 217-28.
Jongwanich J and A. Kohpaiboon. 2008. Export Performance, Foreign Ownership and Trade Policy
Regime: Evidence from Thai Manufacturing, ERD Working Paper No. 140, Economics and
Research Department, Asian Development Bank, 2008.
Kangas K and A. Niskanen. 2003. Trade in Forest Products between European Union and the Central
and Eastern European Access Candidates, Forest Policy and Economics, 5(3): 297–304.
Krugman P.R 1980. Scale Economies, Product Differentiation, and the Pattern of Trade, The
American Economic Review, 70(5), 950-59.
Kumar N. 1994. Determinants of Export Orientation of Foreign Production by U.S. Multinationals: An
Inter-Country Analysis, Journal of International Business Studies, 25(1), 141-56.
Lefebvre E., A.L. Lefebvre and M. Bourgault. 1998. R&D-Related Capabilities as Determinants of
Export Performance, Small Business Economics, 10, 365-77.
Leontief W. 1953. Domestic Production and Foreign Trade: The American Capital Position Re-
Examined, Proceeding of the American Philosophical Society, 97(4), 332-49.
Linder S.B. 1961. An Essay on Trade and Transformation, John Wiley, New York.
Linnemann H. 1966. An Econometric Study of International Trade Flows”, North Holland Publishing
Company, Amsterdam, 1966.
Liu X and C. Shu. 2003. Determinants of Export Performance: Evidence from Chinese Industries,
Economics of Planning, 36, 45-67.
Loecker J. 2007. Do Exports Generate Higher Productivity? Evidence from Slovenia, Journal of
International Economics, 37, 69-98.
Longo R and K. Sekkat. 2004. Economic Obstacles to Expanding Intra-African Trade, World
Development, 32(8), 1309-21.
Majumdar K.S. 1997. The Impact of Size and Age on Firm-Level Performance: Some Evidence
from India, Review of Industrial Organization, 12, 231-41.
133 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Marie M. 2009. Regional Integration and Trade: Controlling for Varying Degrees of Heterogeneity in
the Gravity Model, The World Economy, 32(5), 772-89.
McCallum J. 1995. National Borders Matter: Canada-US. Regional Trade Patterns, American
Economic Review, 85(3), 615-23.
Melitz M.J. 2003. The Impact of Trade on Intra-Industry Reallocations and Aggregate Industry
Productivity, Econometrica, 71(6), 1695-1725.
Michiel V.D. 2002. The Determinants of Export Performance in Developing Countries: The Case
of Indonesian Manufacturing, Eindhoven Centre for Innovation Studies, Working Paper
02.01, Department of Technology Management, Technische Universiteit Eindhoven, The
Netherlands.
Mitze T. 2012. Estimating Gravity Models of International Trade with Correlated Time-Fixed
Regressors: To IV or not IV?, Empirical Modelling in Regional Science, Lecture Notes in
Economics and Mathematical Systems, 667, 165-90.
Mukim M. 2011. Does Exporting Increase Productivity? Evidence from Idia, World Bank Paper
Series.
Muûls M and M. Pisu. 2009. Imports and Exports at the Level of the Firm: Evidence from Belgium, The
World Economy, 32(5), 692-734.
Nguyen B.X. 2010. The Determinants of Vietnamese Export Flows: Static and Dynamic Panel
Gravity Approaches, International Journal of Economics and Finance, 2(4), 122-29
Nguyen H and O. Hiroshi. 2008. Superiority of Exporters and the Causality between Exporting and
Firm Characteristics in Vietnam. The 4th Asia-Pacific Economics Association International
Conference. Beijing.
Nguyen K.D and Y. Heo. 2009. AFTA and Trade Diversion: An Empirical Study for Vietnam and
Singapore, International Area Review, 12(1), 163-92.
Nguyen N.A and T. Nguyen. 2007. Foreign Direct Investment in Vietnam: An Overview and Analysis
the Determinants of Spatial Distribution across Provinces, MPRA Paper No. 1921.
Nguyen T. 2009. Gravity Equation for Different Product Groups: A Study at Product Level.
DEPOCEN Working Paper Series, No. 2009/18.
Nguyen N.A., N. Jones., D.N. Nguyen and D.C. Nguyen. 2009. Capitalising on Innovation for Exports
by the SME Sector. DEPOCEN Working Paper Series, No. 2009/15.
Obben J and P. Magagula. 2003. Firm and Managerial Determinants of the Export Propensity of
Small and Medium-Sized Enterprises in Swaziland, International Small Business Journal,
21(1), 73-91.
Ohlin B. 1933. Interregional and International Trade, Cambridge: Havard University Press.
Özçelic E and E. Taymaz. 2003. Does Innovativeness Matter for International Competitiveness in
Developing Countries? The Case of Turkish Manufacturing Industries, Research Policy,
33, 409-24.
Papadogonas T., F. Voulgaris and G. Agiomirgianakis. 2007. Determinants of Export Behavior in
the Greek Manufacturing Sector, Operational Research, 7(1), 121-35.
134 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Papazoglou C. 2007. Greece Potential Trade Flows: A Gravity Model Approach, International
Atlantic Economic Society, Vol. 13, 403-14.
Papke L.E and J.M. Wooldridge. 1996. Econometric Methods for Fractional Response Variables with an
Application to 401(K) Plan Participation Rates, Journal of Applied Econometrics, 11, 619-32.
Park W.J. 2012. Korea and Vietnam: The National Experience and Foreign Policies of Middle Powers.
Annual Koret Conference.
Porojan A. 2001. Trade Flows and Spatial Effects: The Gravity Model Revisited, Open Economies
Review, 12, 265–80.
Porter M.E. 1990. The Competitive Advantage of Nations, Macmillan, London.
Rahman M. 2003. Panel Data Analysis of Bangladesh’s Trade: The Gravity Model Approach,
University of Sydney, NSW 2006, Australia.
Rankin N., M. Söderbom and F. Teal. 2006. Exporting from Manufacturing Firms in Sub-Saharan
Africa, Journal of African Economies, 15(4), 671-87.
Rault C., R. Sova and A. Sova. 2007. Modeling International Trade Flows between Eastern
European Countries and OECD Countries, IZA Discussion paper No. 2851.
Ricardo D. 1817. On the Principles of Political Economy and Taxation. (London: John Murray,
1821).
Roberts B.A. 2004. A Gravity Study of the Proposed China-ASEAN Free Trade Area, The International
Trade Journal, 18(4), 335-53.
Robson J.A., C.K. Akuetteh, P. Westhead and M. Wright. 2012. Exporting Intensity, Human Capital
and Business Ownership Experience, International Small Business Journal, 30(4), 367-87.
Roper S. and J.H. Love. 2002. Innovation and Export Performance: Evidence from UK and German
Manufacturing Plants, Research Policy, 31, 1087-102.
Roper S., J. H. Love and A. Hīgon. 2006. The Determinants of Export Performance: Evidence for
Manufacturing Plants in Ireland and Northern Irelands, Scottish Journal of Political
Economy, 53(5), 586-615.
Rudolph S. 2010. Computing Multilateral Resistance of Trading Countries, www.freit.org/WP/
TradePatterns.
Samiee S and G.P. Walters. 1990. Influence of Firm Size on Export Planning and Performance,
Journal of Business Research, 20, 235-48.
Sandberg M.H., G.T. Taylor and L.J. Seale. 2002. CARICOM Bilateral Trade: A Preliminary Analysis
Using the Gravity Model, International Agricultural Trade and Policy Center, Institute of Food
and Agricultural Sciences, TPTC. 02-3.
Sapir A. 2001. Trade Regionalism in Europe: Towards an Integrated Approach, ULB Institutional
Repository 2013/8144, ULB - Université Libre de Bruxelles.
Serlenga L. and Y. Shin. 2004. Gravity Models of the Intra-EU Trade: Application of the Hausman-Taylor
Estimation in Heterogeneous Panels with Common Time-specific Factors, Journal of Applied
Econometrics, 22, 361-81.
Shumway T. 2001. Forecasting Bankcruptcy More Accurately: A Simple Hazard Model, Journal
of Business, 74, 101-24.
135 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Silva S.J and S. Tenreyoro. 2006. The Log of Gravity, The Review of Economics and Statistics, 88,
641–58.
Smith A. 1776. An Inquiry into the Nature and Causes of the Wealth of Nations, the University of
Chicago Press.
Sohn C.H. 2005. Does the Gravity Model Explain South-Korea’s Trade Flow?, Japanese Economic Review,
56(4), 417-30.
Sousa C and F. Bradley. 2009. Effects of Export Assistance and Distributor Support on the Performance of
SMEs: The Case of Portuguese Export Ventures, International Small Business Journal, 27(6),
681-701.
Sterlacchini A. 2001. The Determinants of Export Performance: A Firm-Level Study of Italian
Manufacturing, Weltwirtschaftliches Archiv, 137(3), 450-72.
Tapia I., J. Corea and A. Manzanares. 2010. Environmental Strategy and Exports in Medium, Small and
Micro-Enterprises, Journal of World Business, 45, 266-75.
Tinbergen J. 1962. Shaping the World Economy: Suggestions for an International Economic Policy, New York:
Twentieth Century Fund.
Tu T.A and T.M.P. Chu. 2014. On the Border Effect in the Regional Comprehensive Economic
Partnership, A Working Paper funded by the Swiss Secretariat for Economic Affairs (SECO).
Vernon R. 1966. International Investment and International Trade in the Product Life Cycle, The Quarterly
Journal of Economics, 80(2), 190-207.
Vu T.A. 2012. Essays on the Innovation and Intellectual Property System in Vietnam. Université libre
de Bruxelles thesis.
Vu V.H. 2012. Higher Productivity in Exporters: Self-selection, Learning by Exporting or both? Evidence from
Vietnamese Manufacturing SMEs. MPRA Paper No. 40708 August 2012.
Vuong Q.H. 2014. Vietnam's Political Economy: A Discussion on the 1986-2016 Period. CEB Working
Paper Series No 14/010 May 2014. Solvay Brussels School of Economics & Management.
Wagner J. 1995. Export, Firm Size, and the Firm Dynamics, Small Business Economics, 7(1), 29-
39.
Wagner J. 2007. Exports and Productivity: A Survey of the Evidence from Firm-Level Data, The
World Economy, 30(1), 60-82.
Wakelin K. 1997. Innovation and Export Behavior at the Firm Level. Research Policy, 26(1998),
829-41. www.freit.org/WorkingPapers/Papers/TradePatterns/FREIT137.pdf.
Wallace T.D and A. Hussain. 1969. The Case of Error Component Models in Combining Cross-Section
And Time-Series Data, Econometrica, 55-72.
WEF. 2010. The Global Competitiveness Report 2010-2011, available at http://weforum.org/reports.
Zarzoso M.I and F.N. Lehmann. 2003. Augmented gravity model: An Empirical Application to
Mercosur-European Union Trade Flows, Journal of Applied Econometrics, 6(2), 291-316.
Zhang J and G. Kristensen. 1995. A Gravity Model with Variable Coefficients: The EEC Trade
with Third Countries, Geographical Analysis, 27(4), 308-20.
136 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Zhao H and S. Zou. 2002. The Impact of Industry Concentration and Firm Location on Export
Propensity and Intensity: An Empirical Analysis of Chinese manufacturing Firms,
American Marketing Association, 10(1), 52-71.
137 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
APPENDICES
Appendix 1.1. Descriptive statistics for Vietnam’s exports and gravity factors
Variable Mean Std. Dev. Min Max Observations
LnExport
Overall
Between
Within
23.80 3.01
2.81
1.36
15.61
17.91
16.92
31.60
29.98
28.77
N = 1915
n = 170
T-bar = 11.26
LnGDP
Overall
Between
Within
24.60 2.08
2.17
.17
18.91
19.04
23.77
30.20
30.09
25.49
N = 1909
n = 170
T-bar = 11.22
LnGDP/Labor Overall
Between
Within
9.52 1.23
1.24
.12
6.49
6.55
8.88
12.00
11.91
10.25
N = 1873
n = 165
T-bar = 11.35
LnDGDP Overall
Between
Within
2.23 .24
.23
.05
-.37
1.27
.55
2.71
2.70
2.59
N = 1915
n = 170
T-bar = 11.26
LnExchangrate Overall
Between
Within
0.39 0.22
0.15
0.15
-1.50
-0.50
-1.12
2.10
1.16
0.98
N = 1913
n = 170
T-bar = 11.25
LnDistance Overall
Between
Within
8.96 .64
.63
0
6.17
6.17
8.94
9.85
9.85
8.96
N = 1915
n = 170
T-bar = 11.26
ASEAN Overall
Between
Within
.05 .22
.21
0
0
0
.05
1
1
.051
N = 1915
n = 170
T-bar = 11.26
Landlocked Overall
Between
Within
.18 .38
.40
0
0
0
.18
1
1
.18
N = 1915
n = 170
T-bar = 11.26
Border Overall
Between
Within
.19 .13
.13
0
0
0
.19
1
1
.19
N = 1915
n = 170
T-bar = 11.26
Colony Overall
Between
Within
.01 .08
.07
0
0
0
.01
1
1
.01
N = 1915
n = 170
T-bar = 11.26
138 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Appendix 1.2. Results for test for including time dummies
(1) year1998 = 0
(2) year1999 = 0
(3) year2000 = 0
(4) year2001 = 0
(5) year2002 = 0
(6) year2003 = 0
(7) year2004 = 0
(8) year2005 = 0
(9) year2006 = 0
(10) year 2007 = 0
(11) year2008 = 0
(12) year2009 = 0
F (12, 1849) = 49.91
Prob > F = 0.0000
139 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Appendix 2.1. Descriptive statistics for footwear firm’s exports and gravity factors
Variable Mean Std. Dev. Min Max Observations
LnExport
Overall
Between
Within
19.21 2.56
2.44
0.95
8.23
8.77
12.65
27.63
26.83
25.44
N = 23380
n = 9267
T-bar = 2.52
LnGDP
Overall
Between
Within
26.95 1.58
1.72
.03
19.31
19.31
26.71
30.20
30.20
27.18
N = 23265
n = 9237
T-bar = 2.518
LnGDP/Labor Overall
Between
Within
10.60 .76
.82
.03
6.57
6.57
10.31
11.89
11.89
10.95
N = 23237
n = 9220
T-bar = 2.52
LnDGDP Overall
Between
Within
9.65 1.11
1.18
.05
3.44
3.44
8.56
11.21
11.21
10.35
N = 23265
n = 9237
T-bar = 2.51
LnExchangrate Overall
Between
Within
0.07 0.09
0.07
0.07
-.36
-.36
-.26
.61
.61
.46
N = 23380
n = 9267
T-bar = 2.52
LnDistance Overall
Between
Within
8.86 .72
.72
0
6.17
6.17
8.86
9.85
9.85
8.86
N = 23380
n = 9267
T-bar = 2.52
ASEAN Overall
Between
Within
.07 .27
.27
0
0
0
.07
1
1
.07
N = 23380
n = 9267
T-bar = 2.52
Landlocked Overall
Between
Within
.06 .25
.25
0
0
0
.06
1
1
.06
N = 23380
n = 9267
T-bar = 2.52
Border Overall
Between
Within
.02 .16
.18
0
0
0
.02
1
1
.02
N = 23380
n = 9267
T-bar = 2.52
Colony Overall
Between
Within
.03 .17
.17
0
0
0
.03
1
1
.03
N = 23380
n = 9267
T-bar = 2.52
140 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Appendix 2.2. Descriptive statistics for rice firm’s exports and gravity factors
Variable Mean Std. Dev. Min Max Observations
LnExport
Overall
Between
Within
20.76 2.24
2.20
0.70
8.20
8.20
14.93
29.15
28.44
25.15
N = 4019
n = 2574
T-bar = 1.56
LnGDP
Overall
Between
Within
25.61 1.99
2.07
.04
19.45
19.45
25.34
30.20
30.20
25.84
N = 3984
n = 2555
T-bar = 1.55
LnGDP/Labor Overall
Between
Within
9.65 1.23
1.22
.03
6.58
6.58
9.37
11.61
11.61
9.93
N = 3975
n = 2548
T-bar = 1.56
LnDGDP Overall
Between
Within
8.43 1.67
1.66
.33
2.93
2.93
5.12
11.21
11.21
10.69
N = 3972
n = 2544
T-bar = 2.56
LnExchangrate Overall
Between
Within
0.08 0.09
0.08
0.04
-.36
-.36
-.13
.61
.61
.28
N = 4022
n = 2577
T-bar = 1.56
LnDistance Overall
Between
Within
8.73 .75
.75
0
5.37
5.37
8.73
9.85
9.85
8.73
N = 4021
n = 2576
T-bar = 1.56
ASEAN Overall
Between
Within
.14 .35
.31
0
0
0
.14
1
1
.14
N = 4021
n = 2576
T-bar = 1.56
Landlocked Overall
Between
Within
.03 .17
.20
0
0
0
.03
1
1
.03
N = 4021
n = 2576
T-bar = 1.56
Border Overall
Between
Within
.03 .17
.18
0
0
0
.03
1
1
.03
N = 4021
n = 2576
T-bar = 1.56
Colony Overall
Between
Within
.009 .09
.08
0
0
0
.009
1
1
.009
N = 4021
n = 2576
T-bar = 1.56
141 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Appendix 2.3. Descriptive Statistics for wood and wood products firm’s exports and gravity
factors
Variable Mean Std. Dev. Min Max Observations
LnExport
Overall
Between
Within
18.60 2.36
2.16
0.88
3.59
3.59
10.23
27.49
27.49
25.89
N = 35544
n = 19174
T-bar = 1.85
LnGDP
Overall
Between
Within
27.56 1.61
1.70
.03
19.86
19.86
27.25
30.20
30.20
27.83
N = 35251
n = 19038
T-bar = 1.85
LnGDP/Labor Overall
Between
Within
10.74 .72
.79
.02
6.57
6.57
10.44
12.01
12.01
11.08
N = 35214
n = 19004
T-bar = 1.85
LnDGDP Overall
Between
Within
9.89 0.98
1.07
.10
2.32
3.14
7.29
11.21
11.21
12.48
N = 35251
n = 19038
T-bar = 1.85
LnExchangrate Overall
Between
Within
0.08 0.11
0.10
0.06
-.36
-.36
-.26
.61
.61
.44
N = 35555
n = 19184
T-bar = 1.85
LnDistance Overall
Between
Within
8.74 .70
.74
0
6.17
6.17
8.74
9.85
9.85
8.74
N = 35555
n = 19184
T-bar = 1.85
ASEAN Overall
Between
Within
.08 .27
.29
0
0
0
.08
1
1
.08
N = 35555
n = 19184
T-bar = 1.85
Landlocked Overall
Between
Within
.03 .18
.05
0
0
0
.03
1
1
.03
N = 35555
n = 19184
T-bar = 1.85
Border Overall
Between
Within
.06 .24
.25
0
0
0
.06
1
1
.06
N = 35555
n = 19184
T-bar = 1.85
Colony Overall
Between
Within
.04 .21
.19
0
0
0
.04
1
1
.04
N = 35555
n = 19184
T-bar = 1.85
142 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Appendix 3.1. Descriptive statistics for footwear firms
Variables
Observations Mean Std.Dev Min Max
Export intensity 237 .29 .36 1.83e-08 .99
LnWage 237 3.23 .70 .38 5.79
LnEmployee 237 5.57 1.89 .69 9.99
Capintensity 237 485.08 4751.14 -160.46 66349.77
Stateowner 237 .43 .49 0 1
LnFirmage 237 1.93 .74 1 3.95
Location central 237
Location north 237 .16 .37 0 1
Location south 237 .78 .40 0 1
143 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Appendix 3.2. Descriptive statistics for rice firms
Variables
Observations Mean Std.Dev Min Max
Export intensity 83 .21 .24 1.90e-06 .95
LnWage 83 3.55 .88 -.34 5.30
LnEmployee 83 4.44 1.42 1.38 7.98
Capintensity 83 155.94 283.63 -159.24 1604.24
Stateowner 83 .79 .40 0 1
LnFirmage 83 2.02 1.01 0 4.04
Location central 83
Location north 83 .18 .38 0 1
Location south 83 .74 .43 0 1
144 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Appendix 3.3. Descriptive statistics for wood and wood products firms
Variables Observations Mean Std.Dev Min Max
Export intensity 1234 .33 .33 1.04e-07 .99
LnWage 1234 3.22 .64 -.182 7.32
LnEmployee 1234 4.68 1.47 1.09 11.11
Capintensity 1234 72.47 339.58 -3463.64 7954.80
Stateowner 1234 .55 .49 0 1
LnFirmage 1234 1.69 .71 0 3.97
Location central 1234
Location north 1234 .15 .36 0 1
Location south 1234 .65 .47 0 1
145 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Appendix 4.1. Descriptive statistics with quantiles for footwear firms
Variables
n
Mean
S.D
Quantiles
Min .25 Mdn .75 Max
Export
intensity (logit)
237 -2.70 4.22 -17.81 -
5.44
-2.53 0.39 8.75
LnWage 237 3.24 0.70 0.38 2.97 3.26 3.52 5.79
LnEmployee 237 5.58 1.89 0.69 4.16 5.62 7.01 10.00
Capintensity 237 485.08 4751.15 -160.47 -
3.28
8.46 36.89 66349.77
Stateowner 237 0.44 0.50 0.00 0.00 0.00 1.00 1.00
LnFirmage 237 1.93 0.75 0.00 1.39 1.95 2.48 3.95
Appendix 4.2. Descriptive statistics with quantiles for rice firms
Variables
n
Mean
S.D
Quantiles
Min .25 Mdn .75 Max
Export
intensity (logit)
83 -2.33 2.48 -13.18 -3.72 -1.86 -0.68 3.01
LnWage 83 3.56 0.88 -0.34 3.14 3.56 4.06 5.31
LnEmployee 83 4.45 1.43 1.39 3.53 4.48 5.44 7.98
Capintensity 83 155.95 283.64 -159.24 14.41 65.93 174.89 1604.24
Stateowner 83 0.80 0.41 0.00 1.00 1.00 1.00 1.00
LnFirmage 83 2.03 1.00 0.00 1.39 2.08 2.77 4.04
146 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Appendix 4.3. Descriptive statistics with quantiles for wood and wood products firms
Variables
n
Mean
S.D
Quantiles
Min .25 Mdn .75 Max
Export
intensity (logit)
1234 -1.93 3.44 -16.08 -3.94 -1.32 0.74 9.01
LnWage 1234 3.23 0.65 -0.18 2.89 3.23 3.57 7.32
LnEmployee 1234 4.69 1.47 1.10 3.66 4.74 5.68 11.12
Capintensity 1234 72.48 339.58 -3463.6 1.11 21.25 64.55 7954.80
Stateowner 1234 0.56 0.50 0.00 0.00 1.00 1.00 1.00
LnFirmage 1234 1.70 0.72 0.00 1.10 1.79 2.20 3.97
Appendix 4.4. Descriptive statistics with quantiles for the three sectors firms
Variables
n
Mean
S.D
Quantiles
Min .25 Mdn .75 Max
Export
intensity (logit)
1554 -2.06 3.54 -17.81 -4.20 -1.53 0.60 9.01
lnWage 1554 3.25 0.68 -0.34 2.91 3.24 3.58 7.32
lnEmployee 1554 4.81 1.57 0.69 3.71 4.80 5.83 11.12
Capintensity 1554 139.86 1883.6 -3463.64 0.38 19.35 64.17 66349.7
Stateowner 1554 0.55 0.50 0.00 0.00 1.00 1.00 1.00
lnFirm age 1554 1.75 0.75 0.00 1.39 1.79 2.20 4.04
147 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Appendix 5. Logit transformed export intensity for the three sectors
0
.05
.1.1
5
Den
sity
-20 -10 0 10lexportintensity
kernel = epanechnikov, bandwidth = 0.7319
Kernel density estimate
148 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Appendix 6.1. Graph check for normality in the residuals (footwear firms)
Appendix 6.2. Graph check for non-normality in the extremes of the data
(footwear firms)
0
.02
.04
.06
.08
.1
Den
sity
-10 -5 0 5 10 15Residuals
Kernel density estimate
Normal density
kernel = epanechnikov, bandwidth = 1.1587
Kernel density estimate-1
0-5
05
10
Res
idua
ls
-10 -5 0 5 10Inverse Normal
149 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Appendix 7.1. Graph check for normality in the residuals (rice firms)
Appendix 7.2. Graph check for non-normality in the extremes of the data (rice firms)
0
.05
.1.1
5.2
Den
sity
-10 -5 0 5Residuals
Kernel density estimate
Normal density
kernel = epanechnikov, bandwidth = 0.8575
Kernel density estimate
-10
-50
5
Res
idua
ls
-5 0 5Inverse Normal
150 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Appendix 8.1. Graph check for normality in the residuals
(wood and wood products firms)
Appendix 8.2. Graph check for non-normality in the extremes of the data
(wood and wood products firms)
0
.05
.1.1
5
Den
sity
-15 -10 -5 0 5 10Residuals
Kernel density estimate
Normal density
kernel = epanechnikov, bandwidth = 0.6974
Kernel density estimate-1
5-1
0-5
05
10
Res
idua
ls
-10 -5 0 5 10Inverse Normal
151 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Appendix 9.1. Predictions in units of export intensity (footwear firms)
Variables
Observations
Mean
Std. Dev.
Min
Max
Export intensity
237
.29
.36
1.83e-08
.99
Export intensity1 (invlogit)
237
.11
.14
3.85e-09
.98
Appendix 9.2. Predictions in units of export intensity (rice firms)
Variables
Observations
Mean
Std. Dev.
Min
Max
Export intensity
83
.21
.24
1.90e-06
.95
Export intensity1 (invlogit)
83
.11
.075
.01
.37
Appendix 9.3. Predictions in units of export intensity (wood and wood products firms)
Variables
Observations
Mean
Std. Dev.
Min
Max
Export intensity
1234
.33
.33
1.04e-07
.99
Export intensity1 (invlogit)
1234
.16
.13
.001
.94
152 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Appendix 10.1 . Covariate’s effects vary across quantiles (footwear sector)
153 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Appendix 10.2. Covariate’s effects vary across quantiles (rice sector)
154 VU THI HANH – Dissertation – Essays on the Export Performance of Vietnam
Appendix 10.3. Covariate’s effects vary across quantiles (wood and wood products sector)
Recommended