Does economic growth affect environmental quality in some east asian countries

UNIVERSITY OF ECONOMICS INSTITUTE OF SOCIAL STUDIES HO CHI MINH CITY THE HAGUE VIETNAM THE NETHERLANDS VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS DOES INCOME P

UNIVERSITY OF ECONOMICS INSTITUTE OF SOCIAL STUDIES

HO CHI MINH CITY THE HAGUE

VIETNAM THE NETHERLANDS

VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

DOES INCOME PER CAPITA AFFECT

CARBON DIOXIDE EMISSIONS IN SOME

EAST ASIAN COUNTRIES?

BY

PHAM THI THU HUYEN

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

HO CHI MINH CITY, MAY 2012

STUDIES

HO CHI MINH CITY THE HAGUE

VIETNAM THE NETHERLANDS

VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

DOES INCOME PER CAPITA AFFECT

CARBON DIOXIDE EMISSIONS IN SOME

EAST ASIAN COUNTRIES?

A thesis submitted in partial fulfilment of the requirements for the degree of

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

By

PHAM THI THU HUYEN

Academic Supervisor:

PHAM KHANH NAM

HO CHI MINH CITY, MAY 2012

of this thesis From bottom of my heart, I sincerely thank him for all

I grateful acknowledge to Prof., Dr Nguyen Trong Hoai, Dr Nguyen Van Ngai and

Dr Phan Dinh Nguyen – Public Defense Committee for their important comments and explanations that support me to fulfill the thesis

This study also benefits greatly from the enthusiastic assistance of Dr Le Van Chon, whom I got knowledge on econometric and suggested the suitable models for the study

I would like to express my thanks for all students of MDE 16 for their unceasing assistance and encouragement during the course I also show deep gratitude to the lectures, VNP staffs, library staffs for their helps in accumulating knowledge, accessing dataset and materials as well

Last but not least, it gives my deepest grateful to my family members, husband, managers and colleagues for their dear encouragement, give favorable times and opportunities to finish the M.A course as well as the thesis

I pledge to bear full responsibilities for errors, omissions and shortcomings of the study

TABLE OF CONTENTS

CHAPTER 1: INTRODUCTION 1

1.1 Problem Statement 1

1.2 Objectives of study 4

1.3 Research questions 4

1.4 Scope of the research 5

1.5 Structure of the thesis 5

CHAPTER 2: EMPIRICAL AND THEORETICAL BACKGROUND 6

2.1 Theoretical background 6

2.2 Empirical review 9

CHAPTER 3: RESEARCH METHODOLOGY 14

3.1 The conceptual framework 14

3.2 Variables 15

3.2.1 Dependent environment variables 15

3.2.2 Explanatory variables 17

GDP per capita 17

Foreign Direct Investment (FDI) 17

Trade openness 19

Population density 20

3.3 Data……… ……… 22

3.4 The Econometric model 22

CHAPTER 4: DATA ANALYSIS 28

4.1 Carbon Dioxide Emissions and GDP for 6 East Asian countries 28

4.2 Descriptive analysis 30

4.3 The Environment Kuznets curve (EKC) 32

4.4 Determinants of pollution 40

CHAPTER 5: CONCLUSION AND POLICY IMPLICATIONS 46

5.1 Main findings and policy implications 46

5.2 Limitation of the research and recommendations for further study 50

REFERENCES 52

A: Summary of empirical studies on different the EKC hypothesis 57

C: Description of CO2 –GDP of six East Asian countries individually 74

LIST OF FIGURES Figure 2.1 The Kuznets Curve (1955) and the Environmental Kuznets Curve 7

Figure 2.2 The Environmental Kuznets Curve in N-shape 9

Figure 4.2 CO2 emissions and GDP for the period of 1990-2007 29

Figure 5.1 An overview of determinants on environmental quality 48

Table 3.1 Summary of variables description 21

Table 4.1 Summary Statistic on the sample observations 30

Table 4.3 Estimated results for the EKC with different models

(without time trend)

33

Table 4.4 A comparison of estimated results for the EKC with FEM

models

34

Table 4.5 Estimated results for the EKC (N-shape) with FEM 36

Table 4.6 Estimated results for EKC with FEM time effects 38

Table 4.7 Estimated results for FEM with extended model (time

effects)

41

EKC : Environmental Kuznets curve

FDI : Foreign Direct Investment

NGOs : Non-government organizations

CO 2 : Carbon dioxide emissions per capita

GDP : Gross Domestic Products

GHG : Greenhouse gas

OLS : Ordinary least square

FEM : Fixed Effects Model

REM : Random Effects Model

NAFTA : The North American Free Trade Agreement

CHAPTER 1: INTRODUCTION 1.1 Problem Statement

For many centuries, the environmental degradation has become the issue of life with economic development Since the industrial revolution began in 18th century, it was

as a real problem Industrial revolution with the technological progress brought great achievements, especially the worldwide usage of fossil fuels and coal in various industries As a result, global economy has growth fast along with human who significantly consumed natural resources leading environmental quality is worsening seriously (Irina, 2008) This implies that economic growth may affect environmental quality

Every nation wishes to pursue targets of economic growth without damaging environment This is not feasible when history shows that any country becomes an industrial power, in the process of development, causing environmental damages, which cannot solve for a short time In order to serve quick economic growth targets, the policies of incomprehensive industrialization and urbanization with massive exploitation of natural resources have caused environmental problems in these countries for recent years Consequently, the air pollutions and global warming have continuously increased by human’s uncontrolled activities in the development process For China case, environmental degradation is becoming serious problems, which not only affects strongly to domestic but also causes international consequences Some specialists judged that environmental pollution is creating challenges to government It is also a long-term burden on the Chinese people and “China’s problem has become the world’s problem” (Kahn and Yardley,

2007, p.2)

In the current context, environmental problem is not for an individual country to cope with It becomes the common hot issues and needs co-operation of worldwide countries because of its global matter Kim (1996) revealed that due to the quick economic growth, environmental pollution has been in red alarm in China and it

crosses the boundaries to South Korea and Japan Especially, the global pollutants,

in which carbon dioxide emissions are particular pollutant, causing global warming Carbon dioxide emissions have dramatically increased for recent century due to human economic development

Measuring environmental-economic relationship is necessary for any country to orient adequate economic- environmental policy The Environment Kuznets curve (EKC) is reckoned as the best ruler in estimating this relationship since Grossman and Krueger made empirical study on the impact of the economic growth on environment in NAFTA countries in 1991 Under the EKC hypothesis, environmental quality changes with income levels It displays in a bell shape curve Environment degrades with income levels in the first stage of development and then improves as soon as it passes a definite high-income level

Researchers use the EKC hypothesis commonly to connect economic growth with environmental reduction in both single countries and country groups with different models Local pollutants and global pollutants are treated as dependent variables Beside income, other factors are applied as explanatory variables Although the results are not the same for all studies, the economic growth links the environmental deterioration and vice versa However, this theory has still been debated for years because various empirical studies found ambiguous results on the EKC but it has meaning for policy makers to find the best solutions for development sustainability (Dinda, 2004)

In this study, the research concerns if fast economic growth affects the environmental quality in some East Asian countries, specifically six countries consist of three developing countries in Southeast Asia: Malaysia, Thailand, Vietnam and one developing country: China and two developed country in Northeast Asia: Korea and Japan, using cross- country balanced panel data set for the period of 1990-2007

Four developing countries are emerging economies China is now one of the world’s largest economies with World Bank’s ranked by nominal GDP at the third

in 2009, the second position in 2010 and predicts to keep the second rank in 20111 Thailand and Malaysia name in the second generation of newly industrialized economies Vietnam is also thought as the third generation of newly industrialize economy with fast GDP growth in the period of 2000-2009 While Japan and South Korea list the developed countries in East Asian region These countries are not avoidable the environmental damages during the fast economic growth (Nguyen, 2009; Choi et al, 2010)

One of the significant messages of the EKC is that it is not necessary to damage environment during the development and growth if there is suitable policy in technology application Estimating the economic-environmental relationship bases

on an EKC can make important sense in policy implications for six East Asian countries Developing countries like China, Thailand, Malaysia and Vietnam can get lessons and experiences from high income countries such as Japan and South Korea The new emerging country like Vietnam draws more useful lesson in policy establishment from fast developing economies as China, Malaysia and Thailand in term of economic growth and environmental problems solving

This study suggests that there is the existence of an EKC for six East Asian countries It means that income affects environmental damages and it will cure environment when income reach high level Rapid economic growth is vital but not necessary to damage environment Moreover, empirical results shows that beside GDP per capita, other factors contribute environmental pollution Foreign Direct Investment and Trade intensity have positive relation with CO2 emissions but they are not significant statistically Population density is contrary Its coefficient has against sign as expected and has statistically significant at 10 percent level There is

a negative relationship between population density and CO2 emissions per capita It

1 Data from World Development Indicator database, World Bank, 1 July 2011 (Available at http://siteresources.worldbank.org/DATASTATISTICS/Resources/GDP.pdf)

means that population density increases leading to reduce pollution Further study with other pollutants and control variables are suggested reconfirming the EKC hypothesis and making firm policies for six countries

1.2 Objectives of study

The objectives related to six countries in East Asia (China, Vietnam, Thailand, Malaysia, Japan and South Korea) of the thesis include:

 To examine the existence of an Environmental Kuznets curve (EKC) for

a group of six countries

 To find the role of other factors beside income such as trade openness, foreign direct investment (FDI) and population density in determining environmental quality

 To recommend general policies for sustainable development in term of environment and economy of six East Asian countries

1.3 Research questions

Some questions should be presumed in this thesis as following:

1 Does the Environmental Kuznets curve actually exist in a group of six East Asian countries?

2 Does income (GDP per capita) impact on the environmental quality in six East Asian countries?

3 What is the role of other factors beside income, such as population density, trade openness, foreign direct investment (FDI) affect the environmental quality in six East Asian countries?

4 What are recommendations to develop economy harmony with environmental protection in six East Asian countries?

1.4 Scope of the research

The research focuses on finding impacts of economic growth to environmental quality in six East Asian countries under the EKC hypothesis They are China, Vietnam, Thailand, Malaysia, South Korea and Japan The study only investigates

in the period 1990-2007 because full data set before 1990 for some countries (Vietnam, Thailand and Malaysia) is not updated

1.5 Structure of the thesis

The research consists of five chapters as followings:

The chapter 1 explains reason why the topic of thesis is chosen, main objectives, some major research questions and the scope of the research

Chapter 2 provides an overview on theoretical background of the Environmental Kuznets curve (EKC), which measures the relationship between economic growth and environmental damages A review of empirical studies makes clearly about different results of the EKC and opinions on income-environmental relationship and other determinants of environmental quality

Chapter 3 highlights the research methodology Base on the empirical researches and interests, framework of study is mentioned This chapter also describes the general econometrics model as well as variables and data, in which a brief explanation the reasons for choosing variables included It also outlines the process

to analyze data

Chapter 4 shows the statistic results from adopted models Findings are analyzed and to make clear the objectives and answering the questions in the chapter 1 Results of empirical studies by other researchers are also used to strongly support our findings

Chapter 5 resumes main findings and recommends useful policies for six East Asian countries in strategy of economic growth harmony with environmental protection This chapter also identifies the limitations and implications for further study

CHAPTER 2: THEORETICAL AND EMPIRICAL BACKGROUND

of scientists, economists as well as politicians on environmental problems However, the conference did not have a scientific solution to measure the relationship between economic growth and environmental quality

Until year 1991, the Environment Kuznets curve (EKC) was officially born and regarded as an effective instrument to measure a relationship between changes of environmental quality and income levels The EKC has become standard technical vehicle to set up environmental policy (Grossman and Krueger, 1991) Many empirical studies investigated the existence of EKC because it has extremely important policy meaning Nevertheless, not all time studies resulted and implicated base on the EKC

Over the past decades, whenever examine the relationship of economic growth – environmental changes, almost researchers used the EKC hypothesis Empirical studies suggested that there is an inverse-U-shaped relationship between economic growth, usually measured in terms of per capita income, and the environmental quality which proxies by pollutants Specifically, environmental pollution increases

2 The Limits to Growth is a book, which mentioned the aftermath due to world population grows rapidly

while supply of resources is limited Donella H Meadows, Dennis L Meadows, Jorgen Randers and William

W Behrens III are book’s authors ( Available at http://en.wikipedia.org/wiki/The_Limits_to_Growth)

with a rising of income per capita in the first stage of economic growth process When it reaches a certain income level so called turning point, environmental degradation lessens while per capita income still increases This environmental pattern has called the Environmental Kuznets curve (EKC) due to the similarity with the relationship between per capita income and the level of inequality asserted

by Simon Kuznets (1955)

Figure 2.1 The Kuznets Curve (1955) and the Environmental Kuznets Curve

Source: Yandle et al (2004, p.2-3)

Panayotou (2003) and Poudel et al (2009) explained the existence of the EKC because it rooted from the nature of development process Any developed country generally goes through an initial stage of agriculture, which produced labor-intensive products, then moves to industrial period with capital- intensive products and finally towards to service sectors Following these development steps, the environmental performance changes Agricultural production accompanies to make less pollution but environment is damaged during the industrialization and modernization due to natural resources exploitation and energy consumption severing manufacture As development flies over a new phase, service products are favor; environmental practices are improved and polluted intensive sectors transfer

to less developed countries via foreign investment or import polluted intensive products instead of its production As a result, environmental pollution in this stage declines From above ideas, under the EKC hypothesis, economic growth will be the remedy to environmental problems in the future and it is a long run phenomenon

Dinda (2004) gives supplementary ideas on the EKC existing explanation Another reason originates from human’s preference to environmental quality or income elasticity of environmental quality demand When they reach higher income with higher living standards, the demand on environmental quality is greater because they wish to live in a clean environment and have good health conditions A pressure on environmental regulations and protection as well as a care and willing

to pay to green products are indispensible Consequently, environmental degradation is potential reducing

However, for recent years, many researchers have interested in estimating another shape of the EKC, N-shape, which seems to far from the original form This curve points that with changes of income level, pollution declines initially and then increases It means that the EKC does not maintain in the long-run, the inverted U shape exists in the first stage of the income-environment relationship, a new turning point is established to lead to N-shape

The N-shape EKC gives a new outlook on the economic-environmental relationship, any economic activity can affect environment in negative or positive side Even when income level reaches a certain high level, environmental quality cannot be adjusted to improve automatically as the original EKC hypothesis because economic activities are complex and impact environment in different ways (Shafik and Bandyopadhya, 1992) From this point of view, policymakers understand more the economic-environmental relationship in setting strategy for economic growth and environmental protection

Grossman &Krueger (1991, 1995), Shafik (1994), Friedl & Getzner (2003), Sanglimsuwan (2011), Chuku & Agbai (2011), Panayotou (1997), Selden and Song (1994) and Carvalho (2008) found the N- shape instead of the bell shape in relationship of economic growth- environment quality but not many researchers explained why the EKC exists in N-shape replacing inverted-U shape Opschoor (as cited in de Bruyn et al, 1998) illustrates when technology improvement in resources usage leading to pay expensive cost, income growth further will cause environmental degradation It seems that this explanation has been inadequate but there is still no more clear explanation about the N-shape EKC meanings

Figure 2.2 The Environmental Kuznets Curve in N-shape

Source: Carvalho (2008, p.4)

2.2 Empirical review

In 1991, Grossman and Krueger were pioneers who expressing the concept of the relationship of environmental degradation and economic growth, which known officially as the EKC in a study about the environmental influence on countries of the North American Free Trade Agreement (NAFTA) They used cross-country panel of data on urban air pollution via smoke, SO2 emission and SPM with GDP per capita to measure the relationship between air quality and economic growth of

these economies Data took from GEMS for 1977-1984 & EPA for 1985-1988 He found that pollutions increased with GDP per capita at low levels of income but decreased with GDP per capita at higher levels of income This referred to the existence of EKC

Following this study, a series of researches have been implemented in various places and periods, even several people have studied on the EKC’s shape or function form only (Dinda, 2004; Yandle et al., 2004; Panayotou, 2003; Ciegis et al., 2007) However, the empirical reports give ambiguous results and conclusions Some authors affirm the existence of an EKC and implicate the different solutions

to reduce environmental degradation They investigate both groups of countries and individual ones In cases of country group, there are Grossman & Krueger (1991) for 42 countries in North America (NAFTA), Grossman & Krueger (1995) for 42 developing and developed economies during period 1987-1988, Poudel et al (2009) for 15 Latin American nations, Selden and Song (1994) for 30 countries, Eugenio and Roberto (2009) for 73 countries, Holtz-Eakin & Selden (1995) for 131 countries both developing and developed ones, Sanglimsuwan (2011) for 63 countries, Reichert (2007) for 28 emerging and industrialized nations,ect Meanwhile, several authors also find the EKC existence in individual case, for particular, Friedl and Getzner (2003) for Australia, Chuku and Agbai (2011) for Nigeria, Hung and Shaw (2006) for Taiwan and Managi and Jena (2008) for India

On contrast, some researchers confirm that the EKC is not existent and show a linear line In other words, there is unilaterally increasing or decreasing relationship between income and pollution For example, Omojolaibi (2010) for 3 countries in West Africa, Wen and Cao (2009) for China, Khanna and Plassmann (2004) for America, de Bruyn et al (1998) for 4 countries, Day and Grafton (2001) for Canada, Vincent (1997) for Malaysia, Nguyen (2009) for 6 East Asian countries Some studies doubt or show unclear existence of the EKC among pollutants Egli (2002) tested environment- economy relationship of Germany in the period of 1966-

1998 He pointed that the EKC existed for NOx and NH3 but other pollutants (SO2,

CO, CO2, CH4, PM, NMVOC) were unclear He reported that income changes did not influent on pollution emissions in the short-run but they might follow the EKC hypothesis in the long-run time and follow bell-shape curve correlating to income Although the result is mixed, some evidences show the belief of the EKC hypothesis

Similarly, Shafik and Bandyopadhya (1992) studied 149 countries in 1960-1990 They investigate the effect of income per capita to environmental indicators named

as lack of clean water, lack of urban sanitation, SPM, SO2, annual deforestation, total deforestation, dissolved oxygen in rivers, fecal coliform, municipal waste per capita and CO2 emissions per capita The study concludes that there existed the EKC in some pollutants and others are not Most environmental indicators initially rose with income, except for variables of clean water shortage and urban sanitation, which essentially solved by higher income levels Some variables: municipal waste per capita, dissolved oxygen in rivers and deforestation have fitted with quadratic form, which were out of expectation

Cole et al (1997), Shafik (1994) revealed the same results as the two above particular examples Some environmental variables follow the inverted-U-shape curve or N-shape curve but some pollutants did not follow the EKC hypothesis Anyway, there is a little difference in some cases Galeotti et al (2006) measured relationship between CO2 emissions and GDP per capita in OECD countries and non-OECD He reckoned that inverted-U- pattern only was in OECD countries with turning point of $15,000 but not for non-OECD countries Abdubai and Ramcke (2009) found the EKC existence in high-income countries, not for low-income economies when tested the influence of economic growth on the environmental quality, using CFCs, BOD and energy usage relative to per capita GDP

To estimate the relationship between income change and environmental quality, dependent variables are pollutants: CO, CO2, NO2, O3, SO2, TSP, Dioxin, Fecal

coliform, Dissolved oxygen, PM10 or NOx, SPM are proxies for waste materials polluting air, water or soil Researchers use many pollutants or just only one pollutant in their studies because they want a confirmation of the EKC among pollutants Moreover, finding another factors affect environmental quality beside income are necessary to further policy recommendations Thus, they add more explanatory variables

Selden and Song (1994) looked at changes of SPM, CO, NOx and SO2 per capita versus GDP per capita and population density De Bruyn et al (1998) specified for

CO2, NOx and SO2 in relation with GDP per capita, structural change and technological change Wen and Cao (2009) gave the standards model for China with testing relationship between environmental variables such as emission of waste water, waste water per capita, emission of waste gas, waste gas emission per capita, emission of SO2, SO2 emission per capita, emission of solid waste and solid waste emission per capita with national income per capita For case of Hung and Shaw (2006), dependent variables such as CO, SO2, NO2 and PM10 and explanatory variables as population density, number of specialists of air quality protection, air pollution inspection rate, man-made capital stock per capita, employment, education, government expenditure per capita beside income per capita were practiced in his model

Many authors observe environment –economy relationship by using only one pollutant, in which CO2 emissions are supposed as popular environmental indicator with many independent variables in almost investigations For particular, Reichert (2007) with four independent variables as real GDP per capita, percentage of total population live in urban, total commercial energy usage and total power electric consumption to measure the relationship economic growth and environment pollution in 28 worldwide countries in period of 1975-1998 Heil and Selden (2001) investigated with explanatory variables, namely GDP per capita and population in

135 developing and developed countries in years of 1951-1992 Friedl and Getzner (2003) explored the relationship between CO2 emissions as the environmental

quality indicator and GDP per capita, import and value added in service sector as explanatory variables in period of 1960-1999 He and Richard (2010) examined the EKC for Canada, CO2 emissions per capita related to oil price, GDP per capita, industrial structural changes and international trade

Studies investigate the relationship between income and environmental quality, having various results and depending on the environmental indicators A summary

of empirical studies on the difference of the EKC hypothesis is shown in appendix

A The common point of almost studies found that environmental quality is spoilt in

an early stage of economic development and improved in a later stage as economy develops (Dinda, 2004) Following the EKC hypothesis, income level affects environmental quality There is a positive relationship between environmental quality and economic growth at the first stages and a negative side at the later stages

as income gains a certain high level

Summary

This chapter gives an overview on theoretical background, which measures relationship of economic growth –environmental quality The Environmental Kuznets curve (EKC) is mentioned as a foundation of this relationship A basic general model of the EKC also includes Following the EKC hypothesis, it shows an overview of literature review resume with various different results In empirical studies concluded that the EKC is vague but it is meaningful for policy implications Beside the traditional relation of income-environment, other exogenous factors also reconsider A resume of some researches is listed in the appendix A

CHAPTER 3: RESEARCH METHODOLOGY 3.1 The conceptual framework

In this research, the EKC hypothesis is applied to measure the relationship between income level changes and environmental damages Carbon dioxide emissions (CO2) use as pollutant and GDP per capita acts as national income CO2 can partly represent global pollutants Its change means change of environmental quality Moreover, some variables such as population density (Selden and Song, 1994; Abdulai and Ramcke, 2009; Holtz-Eakin and Selden, 1995; Panayotou, 1997; Khanna and Plassmann, 2004; Sanglimsuwan, 2011; Managi and Jena, 2007; Nguyen, 2009; Poudel et al., 2009; Grossman and Krueger, 1991), trade openness (Grossman and Krueger, 1991, 1995; Abdulai and Ramcke, 2009; Panayotou, 2003; Cole et al., 1997; Harbaugh et al., 2001, He and Richard, 2010; Dinda, 2004) and foreign direct investment (Yanchun.Y, 2011; Dinda, 2004; Nguyen, 2009; Shafik and Bandyopadhya, 1992; Abdulai and Ramcke, 2009) are concluded in the model

to explain factors affect environmental quality beside income These factors may cause environmental degradation The reasons why these variables are chosen, mentioning specifically in the part 3.2 of this chapter (data and variables) The impact on environmental quality of the research is shown in conceptual framework

as following:

Figure 3.1 Conceptual framework of the study

EKC Hypothesis Economic Growth

emissions per capita)

3.2 Variables

3.2.1 Dependent environment variables

Many studies have only chosen CO2 as environmental variable to measure the relationship between economic growth and environmental quality in individual or group nations They are Poudel et al (2009), Holtz-Eakin and Selden (1995), Eugenio F & Roberto C (2009), Panayotou (1997), Sanglimsuwan (2011), Friedl and Getzner (2003), He and Richard (2010), Omojolaibi (2010), Moomaw and Unruh (1997), Reichert (2007), Galeotti et al (2006), Heil and Selden (2001) and Chuku and Agbai (2011) In this study, carbon dioxide emissions per capita is chosen as environmental indicator (hereafter called CO2), for some following reasons

Firstly, it is believed that CO2 emissions may one of the primary elements causing greenhouse gas (GHG), which responsible for the problem of global warming Baumert et al (2005) show the persuaded evidence that the global warming and climate change due to GHG emissions, which derive from human activities in burning fossil fuel It increasingly grows because of its damages and dangers to widen environmental scope and human In a report of Intergovernmental Panel on Climate Change, Metz et al (2007) state the truth of human activities in contribution of the global warming over the past 50 years is amazing It covered around at least 90 percent of the observed total The report found that since the mid-

20th century, human activities related to emissions of heat-trapping gases brought almost the increase of averaged temperatures in global level

Friedl and Getzner (2003) concluded that CO2 emissions cause environmental problems and create the sphere of global influence and it takes unimaginable the social costs to solve the problems of global warming as it is accumulated across nations and time Moomaw and Unruh (1997) reconfirmed the potential danger of

CO2 emissions on the global environmental quality, not only for local ones Meanwhile, Arrow et al (1995) mentioned that people think environment is free of

charge For this free rider behavior, human emits CO2 without caring It might lead

to a close association of CO2 emissions and income levels Galeotti et al (2006) also judge the fundamental role of CO2 emissions to GHG and measuring income-

CO2 emissions should be interested specially

Secondly, human activities include the burning of fossil fuel (gas, coal and oil), deforestation, land over-usage, energy consumption originates by transport, facility related to industrial output contribute mainly to CO2 emissions It concentrates in the atmosphere, affecting environment CO2 emissions have been grown significantly in the past centuries and ever increasing in the future both high-income and low-income nations (Heil and Selden, 2001) All above human’s activities seem

to link directly with economic growth and development Moomaw and Unruh (1997) know well the importance of fossil fuel in the development of an economy and claim the CO2 emission for this usage To forecast the CO2 emissions, they predict the demand of energy consumption, which suggest policy of energy and environment

In the fact that a fast economic growth needs higher amount of energy consumption, thus an increase of CO2 and other pollution emissions are inevitable Six countries are chosen in East Asian region for this study because these countries are appreciated in taking large energy consumption Obviously, for these countries, energy plays a vital role in the strategy of the rapid growth It is average estimated about 61% energy consumption of Asia and 20% of world in period 1990-2007 (EIA data, 2011) China, Japan, South Korea and Thailand rank in top 25 CO2

emitting countries, covering 22% of world in 2000 (Baumert et al, 2005, p.111) Some countries as China and Japan have CO2 emissions much higher than other ones To avoid the outlier problem, which can bring great different results in the sample, CO2 emissions per capita is used as a dependent variable in this study

A form of quadratic function allows the examination of an inverted-U shape, which shows relationship between the dependent variable and independent variables This study also wants to check the shape of N-shape but not main task However, cubic function form is not good solution in this study because it may not be appropriate for newly industrialized and developing countries (China, Thailand, Vietnam and Malaysia)

Foreign Direct Investment (FDI)

According to many critics’ opinion, rapid economic growth of developing countries greatly thank for open policy, which attract foreign direct investment and stimulation of trade exchange However, it takes adverse results on living environment and ecosystem Environment seems to unceasing degradation because

of the more industrial pollutions It is right under environmental theory when through investment and international trade, the rich countries tend to import polluted intensive products and find the legal way to move their polluting sectors to

poor countries As mentioned in the previous chapter, high-income people require the clean environment and make pressure on the government on strict environmental regulations As a result, rich countries impose stricter environmental standards and industries related to environment will pay higher cost and lose the advantages To keep profit, they will move these industries to other countries, where environmental regulations are loosen for economic development policies

Poor countries are good destinations for polluted intensive production They need significantly amount of physical capital to support economic growth process and they are easy to attend the race of attract foreign investment by competitive advantages, including low environmental standards and natural resources intensive products Therefore, FDI is an ideal channel for rich countries to transfer polluted industries to poor countries The intensive competition of FDI attraction among poor countries tends to make them lower environmental standards and lead to the

“race to the bottom” in environmental standards (Nguyen, 2009, p.10)

Some empirical studies show that the economies with high investment rates and rapid economic growth put greater pressure on natural resources, particularly in terms of pollutions For example, Shafik and Bandyopadhya (1992) found the positive relationship between CO2 emissions and investment The environmental quality appears going down if the economies pursue policy of rapid economic growth and high investment without environmental response The reason explains why CO2 emissions have close link to investment that investing physical capital directly into manufacture, especially in polluted intensive equipments means energy consumption is increasing and involving some kinds of pollutions

From above arguments, FDI comprises to estimate the impact of FDI on environmental performance in the context of trade liberalization FDI expects to get positive relationship with the CO2

Trade intensity (level of trade openness)

Clearly, international trade contributes remarkably to economic growth Trade openness and growth has complicated relationship (Yanikkaya, 2003) However, economic growth is criticized as a factor to affect the environmental performance Hence, there is a link between international trade and environment Under the theory of the international trade, poor countries and rich countries can exchange products, which they take comparative advantages via international trade Because rich nations have advantage in capital- intensive products while poor countries takes advantages in labor-intensive products

Frankel (2009) stated that:

Openness to trade might encourage some countries to specialize in dirtier activities, and to export their products to others with higher environmental standards Under this “pollution havens” hypothesis, globalization has its primary effect on the distribution of pollution across countries, rather than on the overall average (p.2)

It means that trade openness might influence on environmental quality through economic growth Abdulai and Ramcke (2009) also confirmed this statement and further suggested the different impacts on environment among low-income and high-income nations and difference among regions

Several authors such as Abdulai and Ramcke (2009), Frankel (2009), Grossman and Krueger (1991), Friedl & Getzner (2003) concluded that there is a positive relationship between international trade and environmental performance Cole et al (1997) found a contrast result (negative side) Meanwhile, Abdulai and Ramcke (2009) reveal that for high-income countries, international trade helps reducing energy consumption, leading to improve environmental quality but it makes environment more damages in low -income countries

According to Shafik and Bandyopadhyay (1992), related to trade openness, there are three indicators to measure: share of total export and export in GDP, Dollar’s

index of trade orientation and Parallel Market Premium To research on relationship

of environment –trade openness, almost empirical literature uses the first indicator

In this research, share of total import (X) and export (M) of goods and services in GDP is applied to calculate trade intensity This variable uses to estimate trade openness on environmental quality and expect to have positive sign with CO2

Population Density

Population can be a driving force behind the increase of emissions In reality, the higher population density would make the environment become more polluting because the increase of energy consumption as well as other demands relate to environmental services It implies that an increase in population will force the environmental performance degrading Poudel et al (2009) disclosed the impact of the population density to environmental quality but Sanglimsuwan (2011) found nothing

Many researches use population as an independent variable relative with environmental quality (de Bruyn et al., 1998; Heil and Selden, 2001; Holtz-Eakin and Selden, 1995; Egli, 2002) In this research, population density is tested in spite

of population We assume that the populous country likes China can outlier the general standards comparing the left countries Thus, population in share of land areas, measuring in square kilometers, is the best way to avoid this problem

According to the report of World Resources Institute (Baumert et al, 2005), China, Japan, Thailand and Vietnam listed in the top 25 largest population countries of the world Total average population (1980-2010) of six countries in studied sample is around 25.5 % of world (U.S Census Bureau, 2011) Therefore, the population density adds in the model as an independent variable to calculate pressure of population on environmental damages It is used to control pollutions caused by population growth (Abdulai and Ramcke, 2009, p.10)

As above description and analyses, the expected signs and main empirical references for all variables display in the table 3.1

Table 3.1 Summary of variables

Variables Description Expected

sign Main empirical reference

CO2 Carbon dioxide

emissions per capita

Poudel et al (2009), Holtz-Eakin and Selden (1995), Eugenio F & Roberto

C (2009), Panayotou (1997), Sanglimsuwan (2011), Friedl and Getzner (2003)…

GDP_k GDP per capita

+

Grossman and Krueger (1991), Nguyen (2009), Omojolaibi (2010), Moomaw and Unruh (1997), Sanglimsuwan (2011), Galeotti et al (2006), Heil and Selden (2001), Dong

TRADE Trade intensity

+

Grossman and Krueger (1991, 1995), Abdulai and Ramcke (2009), Panayotou (2003), Cole et al.(1997), Harbaugh et al (2001)…

POP Population density

+

Selden and Song (1994); Abdulai and Ramcke (2009); Holtz-Eakin and Selden (1995); Panayotou (1997), Sanglimsuwan (2011), Managi and Jena (2008) …

3.3 Data

The dataset consists of six East Asian countries They are two developed nations (Japan and South Korea) and four developing nations (China, Malaysia, Thailand and Vietnam) in the period of 1990-2007 The balanced panel data from various sources but mainly draws from World Bank’s World Development Indicators (WDI, 2011) The definition and exact calculated measurement present in the Appendix B

CO2 is collected from Carbon Dioxide Information Analysis Center, Environmental Sciences Division, Oak Ridge National Laboratory, Tennessee, the United States of America It is measured in metric ton per capita GDP comes from the World Bank’s World Development Indicators and estimates in thousand US Dollars (constant 2000 prices) FDI measures in current US Dollar and gets from World Bank, Global Development Finance Trade intensity measurement bases on the data set of import, export and GDP from the World Bank national accounts data and OECD National Accounts data Export and import index are calculated in US Dollars (constant 2000 prices) Population density refers to persons/km2 and from Food and Agriculture Organization and World Bank population estimates

3.4 Econometric model

The Environmental Kuznets curve analysis is an econometric methodology that

assumes that environmental quality or pollutant emissions (CO2) correlate with economic growth, per capita gross domestic product (GDP per capita)

The Environmental Kuznets curve (EKC) analysis commonly used in the literature

is to specify the general functional form (Dinda, 2004, pp.440):

Yit = β0 + β1Xit + β2Xit2 + β 3Xit3 + eit (1) Where, i = 1, N, countries; t = 1, t years; Yit means the environmental pressure of country i at the time t; b0 is country specific intercept; Xit is income in country i at the time t; β1, β2, β3 are the parameter in estimation and eit is the random error in normal estimation

From equation (1), various forms of economic - environmental relationship are tested

(i) If β1 >0 and β2= β3=0: linear shape It presents a monotone increasing linear trend, meaning that as the income rises, environmental pressure increases

(ii) If β1 <0 and β2= β3=0: linear shape It presents a monotone decreasing linear trend, indicating as income level increases, environmental pressure reduces (iii) If β1 >0, β2<0 and β3=0: inverted-U-shape When income reaches certain value, environmental pressure decreases as the income rises

(iv) If β1 <0, β2>0 and β3=0: it presents U-shape which do not expect

(v) If β1 >0, β2<0 and β3>0: N-shape In the initial stage, it is the same as inverted-U-shape and then income level continues rising, environmental pollution rises

(vi) If β1 <0, β2<0 and β3>0: reverse N-shape Environmental degradation falls, then rises and finally increases with income level change

(vii) If β1= β2= β3=0: horizontal line It means that income and environmental quality do not have any relationship

In fact, authors modified the general model, which are suitable with their research objectives Other independent variables add the model and form of model can

change Both non-parametric model and parametric procedures are also used to analyze the relationship between income level and environmental quality

From the equation (1) and review of empirical studies, the most general econometric form testing economic-environmental relationship can be expressed as following:

ENV = f (GDP, GDP2, GDP3, Z)

Where,

ENV is the represent the CO2 emissions per capita, plays the role as dependent variable

GDP denotes of the income (GDP per capita)

Z is other explanatory variables, which suppose to impact to the environmental quality beside income

In the study, we hypothesize that group of six countries, namely China, Japan, Thailand, Vietnam, South Korea and Malaysia would follow the EKC hypothesis on economic- environmental relationship in the period of 1990-2007 It infers there is a similar path in environmental degradation with economic growth among countries

As explained the root of the EKC hypothesis in part of the theoretical background (chapter 2), any country experiences various development stages Each stage has its own changes of economic –environmental relationship Developed countries such

as Japan and South Korea went through initial stages with severe environmental deterioration causing by heavy industrial development strategy and they are now in the right side of the EKC with environmental quality improvement Meanwhile, developing nations like Vietnam, China, Thailand and Malaysia are in the first period of development process with currently serious environmental pollution, it implies that they involve the left side of the EKC According to this assumption, as these countries gain a certain income level, environmental degradation of four above developing countries expect to turn down

With testing the EKC, control variables conclude in the model to explain external influences to environmental performance All dependent variable and independent variables have been explained and selected carefully, in which, persuaded reasons express in the part of variables and empirical study review To illustrate the objectives of the study, we will focus on some tasks

Firstly, we will test the existence of an Environmental Kuznets curve (EKC) Because the EKC stands for the U-inverted curve, the quadratic form is the best choice This kind of form is popularly used by researchers as Nguyen (2009), Day and Grafton (2001), Eugenio and Roberto (2009), Cole et al (1997), Omojolaibi (2010), Heil and Selden (2001), Holtz-Eakin and Selden (1995) One of the purposes of study is to test the existence of EKC in the set of cross-country panel data, not mention the shape of EKC Thus, quadratic function is specific as following form for the group of six countries:

CO2it = β0 + β1GDPit + β2GDPit2 + δt+ µi + uit (2) CO2it is the dependent pollutant variable (CO2 emission per capita) of country i, i=1,2,…,6 in year t, t=1,2,…,18; GDPit is income (GDP per capita) of country i at year t µi is country specific effects, which could be fixed or random and it is time-invariant because it is not change over time δt is time effects The reverse U-shape (EKC) only exists as β1 >0 and β 2<0 and the turning point is estimated by (-β1/2 β2) Secondly, we study determinants of pollutants beside income Other explanatory variables add in the model to estimate the impact to environmental performance CO2it = β0 + β1GDPit + β2GDPit2 + β3FDIit + β4TRADEit + β5POPit + δt+ µi + uit (3)

In which, FDI is Foreign Direct Investment, TRADE is trade openness and POP denotes of Population density

Initially, if we use panel data through pooled OLS model, the major problem of this model is that it does not distinguish among the various countries It means that intercepts for all countries could be identical By mass together different countries

at different times could lead to the heterogeneity The error term is possible be correlated with some of the regressors in the model As a result, the estimated coefficients might be inconsistent and biased (Gujarati and Porter, 2009, p.594-595) However, in the fact that each country has its own different stages of development as well as economic growth (income level) It is important to control the nation differences We suggest that intercepts of each individual country effects are different Moreover, we do not know exactly whether differences across countries can affect remarkably to the changes of CO2 emissions per capita Hence, both FEM and Random Effects Model (REM) could be applied because specific country could fix or random and we could overcome the limitation of Pooled OLS regression

Our data set is balanced cross-country panel Empirical studies prove that it is better

to use Fixed Effects Model (FEM) to consider measuring the relationship between economic growth and environmental quality Some authors such as Selden & Song (1994), Poudel et al (2009), Abdulai and Ramcke (2009), Nguyen (2009), Cole et

al (1997), Omojolaibi (2010), Moomaw and Unruh (1997), Sanglimsuwan (2011), Reichert (2007) use FEM to estimate relationship between economic growth and environmental quality We follow the assumption that six East Asian countries follow the EKC hypothesis So, differences of countries which influent on variation

of CO2 emissions should be controlled FEM could the best We are not sure if FEM is correct choice Hausman statistic test is implemented to examine if FEM or REM is fitted with the data

Summary

In this chapter, conceptual framework establishes on the empirical studies and theoretical background of the EKC hypothesis CO2 emissions per capita is chosen for its global influence Beside GDP per capita, population density, FDI and trade openness are explanatory variables The reasons of variable choice and data sources also make clearly A summary of expected sign and empirical reference for all variables express in table 3.1 and detail definition of variables and data sources are

in the appendix B Besides, econometric model is designed This chapter provides

an overview of research methodology In details, Random Effects Models and Fixed Effects Model suggest to apply for balanced panel data set Hausman test for finding the adequate model with data

CHAPTER 4: DATA ANALYSIS 4.1 Carbon Dioxide Emissions and GDP for six East Asian countries

The figure 4.1 below expresses the CO2 emissions per capita from 1990 to 2007 for six countries It reveals that CO2 emissions tend continuously increasing for each country CO2 emissions in developed nations as Japan and Korea are higher than in developing countries In details, CO2 emissions average obtain 9.69 metric ton per capita for Jaspan, 8.59 metric ton per capita for South Korea, 5.5 metric ton per capita for Malaysia, 3.05 metric ton per capita for China, 3.19 metric ton per capita for Thailand and 0.70 metric ton per capita for Vietnam

Figure 4.1 CO 2 emissions for the period of 1990-2007

We cannot deny the economic growth could cause an increase of CO2 emissions

On average of years, GDP per capita is around 36,575 US Dollars for Japan; 10,820

US Dollars for South Korea; 3, 835 US Dollars for Malaysia; 1,995 US Dollars for Thailand; 936 US Dollars for China and 386 US Dollars for Vietnam Japan is the most developed nation with the highest GDP per capita and highest CO2 emissions Vietnam is less developed nation with lowest GDP per capita and less CO2

emissions in the sample

Under the EKC hypothesis, environmental performance changes with economic growth stages To reach high- income level, an economy spends years with flexible

economic strategies In the sample, China, Thailand, Malaysia and Vietnam are in the first stage of economic development, GDP per capita of these nations are increasing but still low and CO2 emissions rose in period 1990-2007 and tend to increase in coming time South Korea and Japan’s economy are steady with high income, CO2-GDP relationship seems to be idle and fell in years 2005-2007

The figure 4.2 shows the relationship between CO2 emissions and GDP per capita for the period of 1990-2007 When GDP per capita rises, CO2 emission per capita escalates For particular example, China’s GDP per capita increases from 392 US Dollars in 1990 to 1,864 US Dollars in 2007, equivalently 376% while CO2

emissions per capita raises 129% (2.16 to 4.96 metric ton) Korea’s GDP per capita increases 119.8% (6,895 US Dollars in 1990 to 15.157 US Dollars in 2009) while

CO2 emissions per capita increases 84.2% (5.63 metric ton in 1990 to 10.3 metric ton in 2007) Even newly emerge developing country likes Vietnam, in 1990, GDP per capita got 226.85 US Dollars and CO2 emissions per capita was 0.32 metric ton

In 2007, GDP per capita reached 617.11 US Dollars (increasing 172%), CO2

emissions per capita was 1.31 metric ton (hovering 304%)

Figure 4.2 CO 2 emissions and GDP for the period of 1990-2007

4.2 Descriptive analysis

There are total 108 observations in the dataset, in which six individual countries with 18 years Thus, data is balanced and no value is missing Within the period 1990-2007, on average, 5.12 metric ton of CO2 emissions per capita emits yearly South Korea discharged 10.38 metric ton CO2 per capita in 2007 and CO2 emissions per capita of Vietnam was 0.32 metric ton, the lowest one in 1991 The mean over sample years of GDP per capita is around 9,091 US Dollars Japan reached max 40,707 US Dollars in 2007 while Vietnam only got 226.86 US Dollars

in 1990 Doi Moi process began in 1986 and Vietnam economy opened door in

1989 This is reason why GDP per capita is lowest in this year It finds that the higher GDP per capita, the higher CO2 emissions per capita is This proves that CO2

emissions increase parallel with income

Table 4.1 Summary Statistic on the sample observations

Trade openness of six countries on average is around 0.90 US Dollars In which, Malaysia is 1.95; Thailand is 1.12; Vietnam is 1.03; South Korea is 0.66; China is 0.46 and Japan is 0.20 US Dollars Trade openness of developed nations is less than developing countries It proves that developing countries try to develop economy through trade openness Import and export turnover are much larger than GDP per capita

The mean of population density is 230 persons per km2 Surprisingly, population density of South Korea is the highest (around 500 persons per km2) in 2007 and Malaysia’s population density is the lowest (55 persons per km2) in 1990 China is considered as the populous nation and large area in the world Population density seems to be lower (132 persons per km2 on yearly average and 141 persons per km2

in 2007) The detail descriptive statistics of sample shows in the table 4.1

Table 4.2 Correlation on the sample observations

CO2 GDP_k GDP_ksq GDP_kcb FDI TRADE POP

Table 4.2 provides the simple correlation among variables The correlation between GDP and GDP square; GDP and GDP cubic are perfect, near equal 1 (0.98; 0.96 respectively) because GDP square and GDP cubic derive from GDP There is rather high correlation between GDP and CO2 emissions per capita (0.80) CO2 emissions per capita and population also relate (0.62) Obviously, among these variables have relationship each other, especially, CO2 emissions per capita has high positive relationship with GDP per capita and Population density The economic growth and

population growth affects directly to environment However, it is interesting to find that CO2 emissions per capita, GDP per capita and population density have negative correlation with trade intensity and FDI The complicated relationships among variables will further examine in next sections

4.3 The Environmental Kuznets Curve (EKC)

Our data is balanced panels with long time series From equation (2), some tests are carried out F-test for overall significance of the regression with null hypothesis that coefficients are simultaneously equal to zero or independent variables (GDP per capita and GDP2 per capita) do not effect to dependent variable (CO2 emissions per capita) Testing results F( 2,105) = 656.76 and F> critical F value at 1 percent level

of significance (p-value=0.000) It means alternative hypothesis is accepted Explanatory variables (GDP per capita and GDP2 per capita) are influent on CO2

emission per capita

F-test for null hypothesis is that all intercepts are equal to zero or no difference among intercepts The results shows F(2,100) equal 142.9 and p-value of 0.000 It suggests a rejection of null hypothesis and the FEM is better than Pooled model Omojolaibi (2010) also supplied the same result after testing pooled model and FEM

Hausman test is also implemented to find the best choice of FEM and REM for the data Supposing the null hypothesis that estimators would be similar or individual effects are random Overall statistically, χ2 (4) = 66.99, greater than critical value at

1 percent level As a result, null hypothesis is strongly rejected It means that estimators are different, FEM is preferred than REM Several empirical studies found the FEM is the best for panel data in measuring relationship between income and environmental performance under the EKC hypothesis After testing the pooled model, FEM and REM in his study for income-environment relationship in some East Asian countries, using panel data, Nguyen (2009), Holtz-Eakin and Selden (1995) concluded that FEM is the best choice