A Heterogeneous Relationships between Urbanization, Energy Consumption, Economic Growth on Environmental Degradation: Panel Study of Malaysia and Selected ASEAN+3 Countries

Journal of Environmental Treatment Techniques

2020, Volume 8, Issue 1, Pages: 573-581

J. Environ. Treat. Tech.

ISSN: 2309-1185

Journal weblink: http://www.jett.dormaj.com

A Heterogeneous Relationships between

Urbanization, Energy Consumption, Economic

Growth on Environmental Degradation: Panel Study

of Malaysia and Selected ASEAN+3 Countries

Ali Umar Ahmad , Suraya Ismail , Aminu Hassan Jakada , Ibrahim Sambo Farouq , Atiku

Abubakar Muhammad , Umar Aliyu Mustapha , Ahmad Tijjani Abdullahi , Aminu Muhammad

Fagge³

Faculty of Business and Management, University Sultan Zainal Abidin (UNISZA), Terengaennu Malaysia

Faculty of Social Management Science, Department of Economics, Bayero University Kano, Kano Nigeria

Faculty of Arts & Social Science, Department of Economics & Development Studies, Federal University Dutse, Jigawa Nigeria

Received: 11/12/2019

Accepted: 01/02/2020

Published: 20/02/2020

Abstract

This paper aims to analyse the association among urbanization, economic growth, energy consumption and environmental degradation

based on estimates in the context of second-generation techniques. A Malaysian economy and selected ASEAN+3 were estimated using

Pesaran (1999) Pooled Mean Grouped (PMG) and a panel dynamic common correlated effects (DCCE) technique pioneered by Pesaran and

Chudik (2015) that measures a model of error correction (EC) which is resilient to cross-sectional dependency and co-integration. Evidence

from the findings shows that the main actors or driving forces leading to a high level of environmental degradation are urbanization, economic

growth, and energy consumption for Malaysia and selected ASEAN+3 nations. Also found was the existence of one-way causality running

from economic growth to environmental degradation. It also indicates another one-way causality running from square of economic growth

to environmental degradation. Whereas, a bidirectional causality is found between urbanization and environmental degradation, as well as a

feedback causality among energy consumption and environmental degradation.

Keywords: Urbanization, environmental degradation, Energy consumption, Economic growth, Heterogeneous panel, Malaysia and Selected

ASEAN+3

Introduction¹

(

40) Stable economic growth with that of the least environmental

damage from the climate is a severe challenge to the modern

world. Academics and decision-makers are curious to find out

which determinants are responsible for environmental damage

The most challenging phenomenon facing humanity in the

1st century is global climate change, which is increasingly

destroying ecosystems. Over the past few decades, global

consensus has shown that carbon dioxide (CO ) emissions are one

(43). Besides, many contributing factors like energy usage, power

usage, economic growth, trade openness, urban growth and

transportation are responsible for environmental degradation

of the primary sources of global warming due to rising energy

consumption (25) and (35). Since after the industrial age, the rate

(20”16”13). Energy usage in the Association of Southeast Asian

of growth in CO emissions has been 2.0 ppm per year and in

017, this crossed the net figure of 410 ppm. This record will hit

Nations (ASEAN) economies has significantly increased as a

result of steady growth in urbanism and industrial development.

ASEAN. The Center for Energy (ACE) forecast a rise in the

energy usage of ASEAN nations by 4.4% in 2030, which is higher

than the global average demand for energy growth of 1.4%.

Nonetheless, a comprehensive greenhouse gas emission study

continues to be restricted to ASEAN plus Three (ASEAN+3)

member states. Numerous researches cantered on estimating the

a current pace of up to 450 ppm within a few short decades (1).

Achieving a high rate of economic growth through industrial

production and technological progress is a primary concern for

newly developed countries, which are reciprocally enhancing

international trade, urban population and financial development.

The recently industrialized nations (NIC) contribute 42% of total

global CO emissions as a result of increased economic growth

Corresponding author: Ali Umar Ahmad, Faculty of Business

and Management, Universiti Sultan Zainal Abidin (UNISZA),

Terengaennu Malaysia. E-mail: umarahali204019@gmail.com.

Journal of Environmental Treatment Techniques

2020, Volume 8, Issue 1, Pages: 573-581

deterioration of the environment, energy usage and economic

growth in ASEAN economies. (23), and (17) analyzed the

interconnections between CO pollution and powerful influences

in ASEAN nations, whereas (6) based on only ASEAN 8. (36)

and (10) evaluated the impact of ASEAN-5 nations on foreign

direct investment, energy consumption and CO pollution. This

linkage among energy consumption – economic growth and

carbon-economic growth when analysing the empirical study (3)

Given that the sincerity of climate change and its negative

environmental impact has been given more considerable attention

to the global community, several studies have focused primarily

on the relationship among energy usage, in specific the use of

research, therefore, aims to bridge the gap in the research by

examining the role of global warming among ASEAN+3 nations,

especially CO emissions. The slow growth to Japan has indeed

been consumed by the rise of Korea, China, and ASEA (30),

which has expanded the percentage of ASEAN+3 in the world

GDP.

The contribution of ASEAN+3 nations in world GDP

exceeded the United States of America (US) by 2.05 percent and

the members of the European Union (EU) by 1.49 percent in

fossil fuels and CO pollution (38). Looking at the researches on

power-growth-environment (PGE), Soytaş et al. find Granger

non-causality for the US economy among economic growth and

greenhouse gases. (2) notes that both power use and emission

growth are driven by economic development in France. Although

(45), (11) and (44) find that economic growth drives carbon

emissions in China, Jalil and Mahmud (21) accept proof that the

EKC hypothesis is correct. While (18) promotes bi-directional

connections, neither Soytas and (38) and (15) consider causal

links between Turkey's variables. For the board containing six

Central American nations, the EKC theory was endorsed by (15).

(28) for the USA, (27) for India (37) highlighted the results

endorsing a strong association among economic growth and

energy usage. When the findings are regarded in research, it is

hard to say that there is agreement on the course of the energy-

economic growth-environment partnership. It can be said, though,

that energy usage massively increases economic growth and

production of coal.

012. However, the ASEAN+3 rate of growth GDP is projected

to rise by 27% by 2018. The ASEAN plus 3 (APT) was

institutionalized at the Third ASEAN+3 Conference in Manila of

999 (Association of Southeast Asian Party, 2014) in a joint

declaration on East Asian Cooperation. The APT Work Plan for

Cooperation 2013-2017 was subsequently adopted at the 14th

meeting of APT Foreign Ministers on 30 June 2013. One outline

of the APT Work Plans is to reinforce environmental and

sustainable development cooperation and address climate change

impacts. The rest of the current paper will be structured as

follows. The next chapter includes a review of past studies,

followed by the paragraph on methods. The following section

discusses the experimental findings, and in the last chapter, the

article concludes with the conclusion.

2.1 Econometrics Estimation

The study used the annual data for the period 1970-2018 for

Malaysia and Identified ASEAN+3 nations. State choice was

based on a few criteria; covering developing and emerging

economies, recognizing economic proximity and rate of

integration, and eventually representing Annex B states and non-

Annex B nations in the Kyoto Protocol. The study's variables

included urbanization (population growth), economic growth

Literature Review

The The study of carbon dioxide emission drivers is not a new

research subject. Since the age of industrial development, global

economies have been reorganized from organic to inorganic,

pushing the consumption of fossil fuels for industrial production

in order to meet population demands (28). Such structural change

raises the use of fossil fuel resulting in climate change and drastic

climate change (1). Environmental damage and climate change

are currently a significant concern for policymakers to remedy a

(

GDP percentage growth), energy consumption (EC, kg of oil

equivalent per capita), GDP square, and environmental

degradation (CO emission per capita metric ton). This study

involves parameters in logarithmic terms. The data is collected

from the Development Indicators of the World Bank (2019).

STATA 15 and EViews 10 program for analysis.

healthier environment by shocking CO emissions. It is time to

.2 Testing slope homogeneity

The second problem in a panel statistical analysis is whether

examine and build the conceptual framework by government

measures to prevent environmental damage. A large number of

studies examined environmental pollution control performers on

or not the parameters of the slope are heterogeneous. A strong null

hypothesis is the causality of the entire panel from one parameter

to another by applying the mutual constraint (14). Moreover, due

to the specific characteristics of the region, the parameter

homogeneity assumption is not capable of capturing

heterogeneity (8).

various dimensions, such as population impact on CO .

The linkage between energy consumption, economic growth,

and carbon emissions is a topic that has become the order of recent

times for quite a period in the literature on energy and climate

economics. This connection can be classified into four factions by

several quantitative results; although the first cohort of results

suggests that energy usage induces growth and a one-way linkage

exists, the second group asserts that energy usage improves as a

result of economic growth, the third group claims that the

causality connection is bidirectional. There is no causal link

between energy usage and economic growth, as per the

conclusions of the last factions. Nevertheless, with the

Environmental Kuznets Curve (EKC) theory in research, the

association between economic development and greenhouse gas

emissions is established. Under this theory, the pollution rates are

very high during the first phase of the country's development, but

after a certain developmental level, the emissions are decreased

and lower due to further economic growth. Studies focus on the

A standard F-test is the most common way of testing the null

hypothesis of the slope homogeneity: 퐻 ∶ 훽 = 훽 푢푝표푛 푎푙푙 ꢀ in

contrast to the hypothesis of heterogeneity: 훽 ≠ 훽 Representing

푖

푗

a fraction of non-zero pair-wise slope for ꢀ ≠ ꢁ. The validity of F

test can be seen in a scenario where the cross-section dimension

(N) is small relatively, and the panel's time dimension (T) is

enormous; the explanatory variables are strictly exogenous, and

the variances of errors are homoscedastic. (41) developed the test

of slope homogeneity on the dispersion of individual slope

estimates from an appropriate pooled estimator by relaxing the

homoscedasticity assumption in the F test. Moreover, both the F

test and the Swamy test require models of panel data where N is

small relative to T (41).

Journal of Environmental Treatment Techniques

2020, Volume 8, Issue 1, Pages: 573-581

퐻 : 훽 < ꢊ ꢀ = 1,2,3, … . , 푁 훽 = 표 ꢀ = 푁 ꢏ 1, 푁 ꢏ

ꢍ

푖

ꢍ

푖

ꢍ

… … … , 푁. ꢃꢑℎ푒 푠푒푟ꢀ푒푠 ꢀ푠 푠ꢑ푎ꢑꢀ표푛푎푟푦)

(6)

.3 Testing the cross-sectional dependency

Cross-section dependency must be tested once proceeding for

As shown in Table 1, because the series likelihood values and

further steps. Otherwise, outcomes may be bias and contradictory

7). Therefore, the presence of cross-section dependence in the

co-integration formula are less than 0.05, H0 hypotheses are

firmly denied and the cross-sectional correlation between these

countries has been determined. It indicates that a significant

change in the series often impacts the others in one of the nations.

So, when decision-makers in these countries set their policy, other

nations ' reforms and other external determinants should be taken

into account. Moreover, as cross-section dependence has been

established, this condition should be assessed when selecting the

unit root and co-integration testing technique. Nevertheless, panel

unit root checks and study of co-integration were also used taking

into account the cross-sectional dependence. Findings in Table 2

demonstrate that the sequence at levels are non-stationary,

although at first differences get to be stationary; they are shown

to be first-order integrated, I (1). In this scenario, it was

established that the co-integration association between these

patterns could be checked as the sets under consideration are

incorporated in the same order.

(

series and the equation of cointegration should be checked before

further studies. The presence of cross-sectional dependence

between countries is tested through the (7) LM test when the time

dimension exceeds the cross-sectional dimension. (8) has

improved this test if the time dimension is smaller than the cross-

section dimension, and the time dimension is larger than the

cross-section continuum. If the average group is zero, this test is

biased, but the average individual is distinct from zero. By

applying the variance and the mean to the test statistics, Pesaran

(

7) modified this variation. Therefore, the bias-adjusted LM test

LMadj) is named. The adjusted statistical form of the LMadj test is

as follows:

퐿푀_ꢂ_푑_푗

ꢅ ꢋꢎꢍ

ꢆ

ꢋ

퐿ꢃ퐿 − 1)

ꢃ푅 − 푇 − 1) 휌̂ 푖푗 − 휋̂ 훾푖푗

휗_훾_푖_푗

ꢈ

푖푗

(

ꢄ ∑ ∑ ꢇ휌̂

ꢉ ~ 퐿ꢃꢊ,1)

ꢃ1)

푖ꢌꢍ 푗ꢌ푖+ꢍ

.5 Wasteland Cointegration Test

In the research, many panel co-integration tests allow CSD

^w^h^e^r^e^휋^̂ 훾푖푗 ^s^t^aⁿ^d^s^a^s^t^h^e^a^v^e^r^a^g^e^,^휗훾푖푗 Stands as the variance. The

test of the statistics to be generated here reveals a typical normal

distribution as asymptotic (8). The null hypothesis of the LMadj is

the absence of cross-sectional dependence.

between the different groups in the panel. In our statistical

estimation, the (32) board co-integration experiment will only be

implemented and adopted. This test does not only provide robust

results in small sample sizes, but it could also be adhered in all

instances, whether or not CSD exists, and can handle both nation-

specific intercept and slope dimensions together with trends.

However, (32) adopts systemic (instead of just residual) dynamics

by easing the assumption that the unit root first-generation panel

experiment typically imposes a common factor constraint. (32)

then suggested four residual-based experiments that could be used

to determine the null hypothesis of non-co-integration. Four of

these measures are table statistics, while the other four are team

statistics that are usually shared. The test mainly analyses whether

or not co-integration happens by evaluating whether there is an

error correction for the diverse panel groups and the panel itself.

.4 Unit root Test (CADF)

In this analysis, because cross-section dependence between

the countries in the panel among the variables used was

established, one of the second-generation unit root tests

developed by (7) was used to evaluate stationary of the variables.

Unit root testing can be carried out in the series forming the panel

in each cross-section unit through CADF. So, it is also possible to

estimate the sequence stationary one by one for the overall panel

and each cross-section. CADF test assumes that each country is

affected differently from time effects and that in T > N and N > T

circumstances, spatial autocorrelation is used. By comparing the

statistical values of this test with the CADF critical table values

of Pesaran, stationary for each country is tested. If the value of

CADF's critical table is higher than the value of CADF's statistics,

the null hypothesis is rejected and only that country's series is

found to be stationary. The statistics of the CADF test are

estimated as follows:

The measurements are based on a simple error correction

model:

푘ꢍ푖

− 훽 푋_푖_,_푡_ꢎ_ꢍꢔ ꢏ ꢕ 푎_ꢍ_푖_푗∆푦_푖_,_푡_ꢎ_푗

ꢏ

0푖 푖,푡ꢎꢍ 푖

푗ꢌꢍ

∆

푦_푖_푡= 훿 ꢏ 훿 ꢓ푦

푖

푘푗푖

푗ꢌ푘ꢈ푖

ꢕ

푎_ꢈ_푖_푗∆푋_푖_,_푡_ꢎ_푗ꢏ 휇_푖_푡

(7)

푌

푖,푡

= ꢃ1 − 휑 )훼 ꢏ 휑 푦

ꢏ 휋 ꢀ =

Where 훿 is the error correction term, which already provides

푖

푖,푡ꢎꢍ

푖,푡

projections of the pace of adjustment to the long-run equilibrium

of the group. Therefore, the panel hypothesis and group are

estimated as:

,2,3, … … … . , 푁 푎푛ꢐ ꢑ = 1,2,3, … … … . . , 푇

(2)

휋_푖_푡= ꢒ 푓 ꢏ 휇

ꢃ3)

푖

푡

푖푡

퐻 : ꢀ훿⁼^ꢊ^ꢑ^ℎ^푒^푟^푒^ꢀ^푠^푛^표^푝^푟^푒^푠^푒^푛^푐^푒^표^푓^푐^표^ꢀ^푛^ꢑ^푒^푔^푟^푎^ꢑ^ꢀ^표^푛^푓^표^푟^∀푖

Here 푓 displays unobservable prevalent influence of each

country, 휇_푖_푡Reveals the error of individual-specific. Equation (2)

and (3), as well as unit root hypothesis, can be given as follows:

0푖

푡

퐻 : 훿

ꢍ

0푖

^ꢊ^ꢑ^ℎ^푒^푟^푒^푒^푥^ꢀ^푠^ꢑ^푎^푐^표^ꢀ^푛^ꢑ^푒^푔^푟^푎^ꢑ^ꢀ^표^푛^푓^표^푟^∀푖

(8)

In the panel study, the alternative hypothesis suggests that

equilibrium change is homogeneous around the various groups.

Rejecting the null hypothesis thus means that there is proof of co-

integration in the board as a whole.

∆

푦_푖_푡= 훿 ꢏ 훽푦

ꢏ 휏 푓 ꢏ 휇

푖푡

ꢀ = 1,2,3, … … . , 푁 푎푛ꢐ ꢑ

ꢃ4)

푖

푖,푡ꢎꢍ

푖

1,2,3, … … . . , 푇

퐻 : 훽 = ꢊ 푢푝표푛 푎푙푙 ꢀ

ꢃ푛표푛 − 푠ꢑ푎ꢑꢀ표푛푎푟ꢀꢑ푦) (5)

푖

Journal of Environmental Treatment Techniques

2020, Volume 8, Issue 1, Pages: 573-581

Table 1: Cross-sectional Dependency Test, Testing of Slope Homogeneity and Unit root test

CIPS Level

CIPS First

Difference

-4.564*

-6.190*

CADF Level

CADF First

Difference

-3.799*

-6.776*

-7.343*

Variables

퐿푁퐶푂2_푖_푡

18.15*

8.52*

6.82*

36.75*

20.50*

0.584

-3.640

-0.643

-1.769

-1.038

-0.906

-3.735*

-3.925*

-4.138*

-1.039

퐿푁퐺퐷푃

푖푡

퐿푁퐺퐷푃^ꢈ

푖푡

퐿푁푈푅퐵_푖_푡

퐿푁퐸퐶_푖_푡

-2.763*

-5.333*

-6.664*

-4.001*

Test of Homogeneity

퐿푀

320.80*

69.41*

3.795*

∗

퐿푀 푎ꢐꢁ

퐿푀 퐶퐷^∗

Table 2: Summary results of heterogeneous cointegration tests

Dependent Variable is CO

Test type

With Trend

p-value

Without Trend

Value

Statistic

Value

p-value

Westerlund

4.860*

0.000

-4.703*

0.000

23.034*

0.009

-22.803*

0.000

12.155*

-27.309*

0.000

-11.508*

-18.215*

0.000

and ** denotes rejection of the null hypothesis of no cointegration at 1% and 5% levels of significance respectively for Westerlund estimates. AIC is,

therefore, used in selecting lag.

Group Statistics:

connections between both the parameters in trend and trend-free

instances. The group means statistic ꢃ퐺휏 푎푛ꢐ Gα) rejects the null

at both 1% and 5% level of significance and both panel statistics

ꢃ푃휏 푎푛ꢐ 푃훼) at the 1%, 5% and 10% level of significance,

respectively.

퐻 : ꢀ훿 = ꢊ ꢑℎ푒푟푒 ꢀ푠 푛표 푝푟푒푠푒푛푐푒 표푓 푐표ꢀ푛ꢑ푒푔푟푎ꢑꢀ표푛 푓표푟 ∀_푖

0푖

퐻 : 훿

ꢍ

0푖

ꢊ ꢑℎ푒푟푒 푒푥ꢀ푠ꢑ 푎 푐표ꢀ푛ꢑ푒푔푟푎ꢑꢀ표푛 푓표푟 푠표푚푒 푔푟표푢푝푠,

푏푢ꢑ 푛표ꢑ 푓표푟 표ꢑℎ푒푟푠

.6 Long-run and Short-run Estimates

The alternative hypothesis in group analysis also assumes that

Report on IPAT identity by Ehrlich and Holdren (1971)

the transition in equilibrium is diverse around the various groups,

and denial of the null hypothesis implies evidence of co-

integration in at least one member of the group.

provides the basis for many work relevant to the detection of

environmental degradation drivers. The well-known

identification indicates that environmental destruction is the result

of population (P), affluence (A), and technology (T) and can be

defined as follows:

In order to determine the vibrant assimilation between

Urbanization, Energy usage, Economic Growth on

Environmental Deterioration in Malaysia and Selected

ASEAN+3 nations, this study includes a time series panel

method, notably (32) error-based tests to measure the long-term

connection between Urbanization, Energy usage and Economic

Growth on Environmental damage. The main advantage of the

operation is to analyse the variables ' pre-movement without any

endogeneity concerns. The tests are premised on (31) suggested

structural instead of residual nuances. (32) show better smaller

sample characteristics with small size biases and high voltage

relative to residual-based co-integration technique in the error

typo-based tests.

퐼 = 푃. 퐴. 푇

(9)

ꢖ

ꢗ

ꢘ

퐼 = 휑푃 . 퐴 . 푇

(10)

After taking the logarithm, Eq. (9) is transformed into Eq. (10)

as follows.

ln 퐼 = ln 휑 ꢏ 휎 ln 푃 ꢏ 휌 ln 퐴 ꢏ 휏 ln 푇 ꢏ 푙푛휀

(11)

where 휑 is the coefficient of the model, 휎, 휌, 휏 are exponentials of

independent variables and STIRPAT model random error term is

represented by 휀. In this research, (7) and the Dynamic Panel

Common Correlated Effects (DCCE) method established by (7),

which calculates an error correction (EC) model, scientifically

evaluate the impacts of population (P), affluence (A), technology

(T) and other actors on environmental damage. The (7)

assumptions that variables are exogenous and require feedback

effects among eigenvalues that can lead to problems of

This study of 10 emerging countries (Malaysia and Selected

ASEAN+3 nations) also suits well. Both experiments are evenly

distributed and can integrate unit-specific short-run trends, unit-

specific pattern and slope factors, and cross-sectional dependence

(32). Table 2 above indicates that all four figures refute the null

hypothesis of no cointegration in both trendy and trend-free

economies in Malaysia and chosen ASEAN+3. This means that

in Malaysia and Selected ASEAN+3 nations, there are long-run

Journal of Environmental Treatment Techniques

2020, Volume 8, Issue 1, Pages: 573-581

′

consistency. (7) tackles three main issues, the first issue being

cross-sectional correlations, which are addressed by taking cross-

sectional averages and lagging cross-sectional averages of

predictive factors on the right-hand side of the equation with

predictor variables. The second problem is parameter variability,

which can be overcome using Eberhardt and Presbitero's (2015)

mean group model. The third problem is nuances that can be

solved by introducing the explanatory determinants lag in the

model. The common correlated effects dynamic panel (DCCE)

method solves all the problems mentioned above and provides

more accurate predictions for the expanded STIRPAT model. The

model in question consists of the preceding equations:

퐼_푖_푡= ꢙ ꢏ ∐ 푙

ꢏ 휌푃 ꢏ 휏퐴 ꢏ 휑푇 ꢏ 휇 퐼푉 ꢏ 푄 ꢏ

푖

푖,푡ꢎꢍ

푖푡

ꢢ

ꢏ ꢕ

ꢚ푖

푖푡

ꢢ

ꢕ

∅_ꢍ_푖_ꢡ

̅̅̅̅̅

ꢏ ꢕ

̅̅̅̅̅̅̅

∅_ꢈ_푖_ꢡ푃퐼

̅̅̅̅̅̅̅

∅_ꢣ_푖_ꢡ퐴퐼 ꢏ

푡ꢎꢡ

ꢡꢌ0

푡ꢎꢡ

ꢡꢌ0

푡ꢎꢡ

ꢡꢌ0

ꢢ

∅_ꢤ_푖_ꢡ 푇̅ ̅ 퐼̅ ̅̅̅̅ ꢏ ꢕ

∅₅_푖_ꢡ 퐼̅ ̅푉̅ ̅퐼 ̅̅̅̅ ꢏ 휀

(19)

푡ꢎꢡ 푖푡

represents the dependent

ꢡꢌ0

푡ꢎꢡ

ꢡꢌ0

̅̅̅̅ ̅̅̅̅̅̅̅ ̅̅̅̅̅̅̅ ̅̅̅̅̅̅̅̅ ̅̅̅̅̅̅̅̅

where 퐼 , 푃퐼 , 퐴퐼 , 푇퐼 , 퐼푉퐼

푡ꢎꢡ

variable cross-sectional average, and y represents cross-sections

averages lags. Dependent variable (퐼_푖_,_푡_ꢎ_ꢍ) included as an

explanatory variable because models with dynamic properties

such as Dynamic Common Correlated Estimator (DCCE) uses lag

of independent variable (7).

However, the PMG estimators illustrate both the pooling

suggested by the drawbacks of homogeneity on the long-run

coefficients and the median through classes used to obtain process

for the analyzed model's other short-run parameters and error

correction coefficients. The Pooled Mean group model, including

the long-term relationship between variables, that obey the

methods of Pesaran et al. (1999):

ꢏ 휌푃 ꢏ 휏퐴 ꢏ 휑푇 ꢏ 휇^′_ꢚ_푖퐼푉 ꢏ 푄_푖_푡

(12)

(13)

퐼_푖_푡= ꢙ ꢏ ∐ 푙

푖

푖,푡ꢎꢍ

푖푡

푄_푖_푡= 푄^′ꢏ ∀_푖_푡

푖ꢛ푡

^퐼^푉푖푡

′

푟_푖_푡= (

ꢄ = ꢙ푟 ꢏ 훽 푙

ꢏ 푄 푔 ꢏ 푀

푖푡

(14)

푖

푖 푖,푡ꢎꢍ

푖

푡

푎_푖_푡

ꢥꢎꢍ

푞ꢎꢍ

∆

퐿푁퐶푂2_푖_푡= 훽 ꢏ ∑ 휕 ∆퐿푁퐶푂2

ꢏ ∑ ꢒ ∆퐿푁퐺퐷푃

ꢍ

푖푗

푖푡ꢎ푗

푖푗

푖푗ꢎꢍ

where ꢀ = 1, ꢑ표 푁; ꢑ = 1 ꢑ표 푇; 푙_푖_푡is a dependent variable that

^r^e^p^r^e^s^eⁿ^t^s^C^O²^e^mⁱ^s^sⁱ^oⁿ^;^푃푖푡 is a measure of the population; Ait

is a measure of affluence; Tit is a measure of technology and 퐼푉_푖_푡

Denotes multiple predictor variables for the extended STRIPAT

system, including energy output, labor productivity, residential

housing, density population, energy mix and market accessibility.

푗ꢌꢍ

ꢛꢎꢍ

ꢍꢌ0

ꢈ

ꢏ ∑ 훿 ∆퐿푁퐺퐷푃

푖푗

푖푗ꢎꢍ

ꢍꢌ0

ꢦꢎꢍ

ꢜꢎꢍ

ꢏ ∑ 휑 ∆퐿푁푈푅퐵

ꢏ ∑ 휔 ∆퐿푁퐸퐶

푖푗

푖푗ꢎꢍ

푖푗

푖푗ꢎꢍ

Also ꢙ represents effects specification to the country which are

푖

ꢍꢌ0

unobserved; 푎_푖_푡consistent parameters contingent on some

possible determinants but not dependent on parameter

ꢏ 휋 퐿푁퐶푂2

ꢏ 휋 퐿푁퐺퐷푃

ꢍ

푖푗ꢎꢍ

ꢈ

푖푗ꢎꢍ

ꢈ

푖푗ꢎꢍ

ꢏ 휋 퐿푁퐺퐷푃

ꢏ 휋 퐿푁푈푅퐵

ꢣ

ꢤ

푖푗ꢎꢍ

dependency. 푔 is the economy specific impacts of time-varying

푡

ꢏ 휋 퐿푁퐸퐶

ꢏ 휇_ꢍ_푖_푡

푖푗ꢎꢍ

determinants that are not noted and also reflect downturns that

impact all the newly industrialized nations with the same extent

on a world level. ∀_푖_푡Represent errors which are not correlated

with regressors. Also ∀_푖_푡exhibit shocks which are common and

unobserved as well, furthermore these errors are also weakly

independent across countries and are also serially correlated. ωi

^ꢏ^휀ꢍ푖푡

ꢃ2ꢊ)

where:

∆

is the first difference operator, and

ꢈ

퐿푁퐶푂2, 퐿푁퐺퐷푃, 퐿푁퐺퐷푃 , 퐿푁푈푅퐵, 푎푛ꢐ 퐿푁퐸퐶 Are the five

variables selected in the study. The constants is 훽 , the short-run

and

ꢍ

long-run

coefficients

the

trends

are

denotes matrix for factor loadings, 훽 represents the vector of

푖

휕 , ꢒ , 훿 휑 , 휗 , 푎푛ꢐ 휔 푎푛ꢐ 휋 , 휋 휋 휋 푎푛ꢐ 휋 ,

푖푗 푖푗 푖푗 푖푗 푖푗

coefficients and 푀_푖_푡is a stationary covariance process regardless

of error 푀_푖_푡. Further, it is assumed that vector consisting of factor

푖푗

ꢍ

ꢈ

ꢣ

ꢤ

respectively. 푝, ꢧ, 푟, 푠, 푎푛ꢐ 푧 represents the maximum lag

^l^eⁿ^g^t^h^,^휀ꢍ푖푡 are error terms,

′

loadings (휕 , 휋 ) and coefficients such as Ω = ꢃ푄 , 휌, 휏, 휑, 휇 _ꢚ_푖)

푖

The table above shows the long-run results of DCCE and

PMG estimates. These results revealed that at a 1% level of

significance, a unit increase in GDP leads to 16% and a 29% rise

follow the below models with random coefficients.

푄 = 푧 ꢏ 휃₀_,_푖_,휗₀_,_푖~퐼퐼퐷ꢃꢊ, 훿_ꢜ)

(15)

푖

in CO emissions, respectively. While at a 1% level of

significance, a 1 unit rise in GDP will bring about a 69% decrease

in CO emissions, and subsequently, at a 5% level of significance,

푣푒푐ꢃ푧 ) = 푣푒푐ꢃ푧) ꢏ 휃 휗 ~퐼퐼퐷ꢃꢊ, 훿 )

(16)

(17)

푖

ꢜ,푖, ꢜ,푖

ꢜ

a 1-unit increase in GDP will result in 24% decrease in CO

emissions. Whereas, at a 1% level of

Ω = Ω ꢏ 휃_Ω_,_푖_,휗_Ω_,_푖~퐼퐼퐷ꢃꢊ, 훿_Ω)

푖

significance, a 1-unit increase in POP will leads to 36% and 31%

decline in CO emission, respectively. However, at a 5%

significance level, a 1-unit increase in EU leads to 21% and a 35%

increase in CO emission.

Meanwhile, in the short-run, at a 1% significance level, a unit

increase in GDP leads to 34% and a 13% increase in CO

ꢍ

ꢕ^_^_^_^_푀퐺 =

ꢕ

ꢝ

푖ꢌꢍ

)ꢃ Ωꢞ − Ω )^′

ꢃ Ωꢞ − Ω

(18)

푖

ꢟꢠ

푖

ꢟꢠ

ꢝꢎꢍ

We embraced the method of dynamic joint-related effect

assessment (7) for the measurement equation used as cross-

sectional averages dependent variable greenhouse gases. The

model in Eq. (18) also includes lag of the dependent variable as a

proxy for common factors effect along with P, A, and T and

extended variables.

emission. Also, a unit change in GDP , 27% and 48% increase in

CO emission will result in restively at a 1% level of significance.

While at a 5% level of significance, a unit change in POP will

result in a 35% decrease in CO emission. Whereas, at a 5% level

of significance, a unit increase in the EU will lead to a 21% and

5% increase in CO emission, respectively.

Journal of Environmental Treatment Techniques

2020, Volume 8, Issue 1, Pages: 573-581

Table 3: Estimates Results

Dependent variable: LNCO2it

DCCE

PMG

Variables

Coefficients

Standard

error

value

Coefficients

Standard

error

value

Long-run estimates

LNGDPit

Long-run estimates

0.299*

0.071

0.167

0.097

0.249

0.000

0.002

0.163*

[-3.805]

0.043

0.087

0.249

0.139

0.000

0.046

0.002

0.000

[4.220]

ꢈ

푖푡

LN퐺퐷푃

-0.698*

-4.179]

-0.248**

[-2.640]

[

LNURBit

LNEUit

-0.369*

-3.806]

-0.590*

[2.861]

[

-0.801*

-3.221]

-1.192*

[-8.576]

[

Short-run estimates

ΔLNGDPit

0.347*

3.343]

0.104

0.102

0.174

0.088

0.071

0.000

0.008

0.035

0.005

0.000

0.137*

[2.322]

0.059

0.156

0.069

0.139

0.179

0.006

0.000

0.012

0.002

[

ꢈ

푖푡

ΔLN퐺퐷푃

0.274*

2.686]

0.487*

[3.122]

[

ΔLNURBit

ΔLNEUit

ectt-1

-0.369**

-2.119]

-0.315*

[-4.565]

[

0.210**

2.386]

0.350**

[2.52]

[

-0.438*

-6.147]

-0.557*

[-3.115]

[

Observation

Cross-section

F(P)

598

299.15(0.000)

0.68

0.57

Adj-R

CD Statistics

Optimal lag length

4.77(0.000)

(1,1,1,1,1)

causality of Dumitrescu and Hurlin Granger are:

The results are in line with EKC, as the combination of

economic development and greenhouse gas emissions is

developed with the Environmental Kuznets Curve (EKC) concept

study. Under this hypothesis, pollution rates are very high in the

first stage of development of the country, but after a certain level

of development, emissions are lowered and lower due to more

economic growth, just like the findings of the current study.

퐾

ꢃ푘)

∆퐿푁퐶푂2_푖_,_푡= 훽 ꢏ ∑ 휕_푖∆퐿푁퐶푂2_푖_,_푡_ꢎ_푘

푖

푘ꢌꢍ

퐾

ꢃ푘)

ꢏ ∑ ꢒ_푖∆퐿푁퐺퐷푃푖,푡ꢎ푘

푘ꢌꢍ

퐾

ꢃ푘)

^ꢏ^∑^ꢒ푖 ∆퐿푁퐺퐷푃^ꢈ

.7 Dumitrescu and Hurlin heterogeneous Panel Granger

푖,푡ꢎ푘

Causality Estimates

푘ꢌꢍ

퐾

In heterogeneous panel data models with defined coefficients,

ꢃ푘)

^ꢏ^∑^훿푖 ^∆^퐿^푁^푈^푅^퐵푖,푡ꢎ푘

(

14) developed the non-causality test. The rise in generic causality

checks for panel data suggests evaluating cross-sectional

longitudinal constraints on design coefficients in the context of a

linear autoregressive information generation process. The use of

cross-sectional data will expand the causality data set from one

parameter given to another. The composite board figures of

푘ꢌꢍ

퐾

ꢃ푘)

^ꢏ^∑^휃푖 ^∆^퐿^푁^퐸^퐶푖,푡ꢎ푘

푘ꢌꢍ

ꢏ 휀_푖_,_푡

ꢃ21)

Journal of Environmental Treatment Techniques

2020, Volume 8, Issue 1, Pages: 573-581

The model uses a fixed special effect and a fixed coefficient

model. The heterogeneous no-causality hypothesis the null

∆

퐿푁퐺퐷푃

푖,푡

퐾

ꢃ푘) ꢃ푘) ꢃ푘)

ꢃ푘)

hypotheses: (퐻 :휕_푖, 훿_푖, ꢒ_푖푎푛ꢐ 휃_푖= ꢊ∀_푖_푗= 1, . . 푁). The

ꢃ푘)

푖

ꢃ푘)

^훽푖 ^ꢏ^∑^ꢒ

^∆^퐿^푁^퐺^퐷^푃푖,푡ꢎ푘 ^ꢏ^∑^휕푖 ^∆^퐿^푁^퐶^푂²푖,푡ꢎ푘

value of F-statistics and p-value, which indicates whether to reject

or not to reject the null hypothesis, documents no, or the existence

of causality.

Table 4 below shows the causality relationship between the

determinants concerning CO emissions, which highlights that

푘ꢌꢍ

퐾

ꢃ푘)

^ꢏ^∑^ꢒ푖 ∆퐿푁퐺퐷푃^ꢈ

푖,푡ꢎ푘

^ꢏ^∑^훿푖 ^∆^퐿^푁^푈^푅^퐵푖,푡ꢎ푘

푘ꢌꢍ

퐾

푘ꢌꢍ

there is an existence of one-way causality running from economic

ꢃ푘)

^ꢏ^∑^휃푖 ^∆^퐿^푁^퐸^퐶푖,푡ꢎ푘

growth to CO₂emission. It also indicates another one-way

푘ꢌꢍ

causality running from GDP to CO

emission. Whereas, a

ꢏ 휀_푖_,_푡

ꢃ 22)

bidirectional causality is found between urbanization and CO

emission, as well as a feedback causality among energy used and

CO emission. This is very consistent with the findings of (18)

∆

퐿푁퐺퐷푃^ꢈ

푖,푡

퐾

While (11), (12) and (25) found economic growth driving carbon

emissions (1) accept evidence that the EKC hypothesis is right.

ꢃ푘)

푖

ꢃ푘)

^훽푖 ^ꢏ^∑^ꢒ ∆퐿푁퐺퐷푃^ꢈ

^ꢏ^∑^ꢒ푖 ^∆^퐿^푁^퐺^퐷^푃푖,푡ꢎ푘

푖,푡ꢎ푘

푘ꢌꢍ

퐾

푘ꢌꢍ

퐾

Table 4: Granger causality results

ꢃ푘)

^ꢏ^∑^휕푖 ^∆^퐿^푁^퐶^푂²푖,푡ꢎ푘 ^ꢏ^∑^훿푖 ^∆^퐿^푁^푈^푅^퐵푖,푡ꢎ푘

LNGDP−/

LNCO2

LNGDP←/

−LNCO2

푘ꢌꢍ

퐾

푘ꢌꢍ

→

W^Hⁿ^c

1.703

ꢃ푘)

ꢏ ∑ 휃_푖∆퐿푁퐸퐶_푖_,_푡_ꢎ_푘

.078*

.778

푘ꢌꢍ

ꢫꢬꢭ

푍

1.546

ꢏ 휀_푖_,_푡

ꢃ 23)

ꢝꢪ

ꢫꢝꢮ

푍

ꢝ

퐾

LNGDP −/

LNGDP ←/

−LNCO2

ꢃ푘)

LNCO2

∆

퐿푁푈푅퐵_푖_,_푡= 훽 ꢏ ∑ 훿_푖∆퐿푁푈푅퐵_푖_,_푡_ꢎ_푘

푖

W^Hⁿ^c

1.227

푘ꢌꢍ

퐾

.383*

.494

ꢃ푘)

ꢏ ∑ ꢒ_푖∆퐿푁퐺퐷푃^ꢈ

푍

ꢫꢬꢭ

0.424

푖,푡ꢎ푘

ꢝꢪ

푘ꢌꢍ

퐾

ꢫꢝꢮ

푍

ꢝ

ꢃ푘)

^ꢏ^∑^ꢒ푖 ^∆^퐿^푁^퐺^퐷^푃푖,푡ꢎ푘

LNURB−/

LNCO2

LNURB←/

푘ꢌꢍ

퐾

−LNCO2

2.259*

2.853

W^Hⁿ^c

ꢃ푘)

ꢏ ∑ 휕_푖∆퐿푁퐶푂2_푖_,_푡_ꢎ_푘

.560*

푘ꢌꢍ

퐾

ꢫꢬꢭ

푍ꢝꢪ

ꢃ푘)

15.318

^ꢏ^∑^휃푖 ^∆^퐿^푁^퐸^퐶푖,푡ꢎ푘

ꢫꢝꢮ

푍

ꢝ

푘ꢌꢍ

LNEU−/

LNCO2

LNEU←/

−LNCO2

ꢏ 휀_푖_,_푡

ꢃ24)

→

퐾

_WHnc

ꢃ푘)

∆

퐿푁퐸퐶_푖_,_푡= 훽 ꢏ ∑ 휃_푖∆퐿푁퐸퐶_푖_,_푡_ꢎ_푘ꢏ ∑ 훿_푖∆퐿푁푈푅퐵_푖_,_푡_ꢎ_푘

2.447*

0.337

3.292*

푖

ꢫꢬꢭ

푍

푘ꢌꢍ

ꢝꢪ

퐾

92.398

ꢃ푘)

ꢫꢝꢮ

ꢝ

ꢏ ∑ ꢒ_푖∆퐿푁퐺퐷푃^ꢈ

푍

푖,푡ꢎ푘

푘ꢌꢍ

퐾

ꢃ푘)

ꢏ ∑ ꢒ_푖∆퐿푁퐺퐷푃_푖_,_푡_ꢎ_푘

3 Results and policy implications

This research used the sophisticated panel DCCE approach to

analyse the expanded STIRPAT model for a panel of Malaysia

and Identified ASEAN+3 economies from 1970 to 2018. The

presence of cross-sectional dependence between the nations in the

panel, in other phrases the hypothesis that a macroeconomic

downturn in one of the nations examined might affect others, also

analyzed using the CDLM test formulated by (8) and those whose

variance was rectified by Pesaran, Ullah, and Yamagata and it was

defined that cross-section existed In the co-integration equation

and between the examined countries ' sequence of urbanization,

푘ꢌꢍ

퐾

ꢃ푘)

ꢏ ∑ 휕_푖∆퐿푁퐶푂2_푖_,_푡_ꢎ_푘

푘ꢌꢍ

ꢏ 휀_푖_,_푡

ꢃ2ꢨ)

Where 훽 Remain steady in the dimension of the time, and ꢩ

denotes steady lag orders for all cross-sections of the panel. This

푖

ꢃ푘)

allows ꢒ_푖, 휕_푖, 훿_푖푎푛ꢐ 휃_푖

parameters and coefficients of slope to differ across the groups.

As an autoregressive

Journal of Environmental Treatment Techniques

2020, Volume 8, Issue 1, Pages: 573-581

economic growth, energy usage and environmental pollution.

CADF test designed by (7) examined the presence of unit root in

series in this analysis, taking into account the cross-section

dependence in series, and it was noticed that series were not level

consistent and stable when their first differences were taken.

In this case, the precondition for studying the co-integration

linkage between series was determined. The presence of the co-

integration relation between the series was tested by the test

established by (32) and the cross-section dependence was

considered and the co-integration connection between the series

was observed. Then the calculation of the board is performed

using the DCCE method. Co-integration analysis is also used for

robustness to verify the co-integration relationship between

variables. Also, compared to PMG techniques are the results of

the DCCE method. Finally, the research applied causality checks

for Dumitrescu and Hurlin to explore the causalities of course.

The research to investigate essential factors that are the primary

in an enormous cost decline in renewable energy. The decision-

makers in Malaysia and selected ASEAN+3 nations should take

some steps to personalize the need for energy and encourage the

use of renewable energy at the commercial and industrial level by

creating policies such as user tax subsidies. One of the effective

measures to limit environmental damage in Malaysia and selected

ASEAN+3 countries may be encouraging productive industrial

and household power use.

References

Ahmed, K., Akhondzada, A., Kurnitski, J., & Olesen, B. (2017).

Occupancy schedules for energy simulation in new prEN16798-1 and

ISO/FDIS 17772-1 standards. Sustainable cities and society, 35, 134-

144.

Ang, J. B. (2007). CO2 emissions, energy consumption, and output

in France. Energy policy, 35(10), 4772-4778.

Apergis, N., & Payne, J. E. (2009). CO2 emissions, energy usage, and

output in Central America. Energy Policy, 37(8), 3282-3286.

Apergis, N., & Payne, J. E. (2010). Renewable energy consumption

and growth in Eurasia. Energy Economics, 32(6), 1392-1397.

Beigel, J. H., Tebas, P., Elie-Turenne, M. C., Bajwa, E., Bell, T. E.,

Cairns, C. B., ... & Luke, T. (2017). Immune plasma for the treatment

of severe influenza: an open-label, multicentre, phase 2 randomised

study. The Lancet Respiratory Medicine, 5(6), 500-511.

Borhan, H., Ahmed, E. M., & Hitam, M. (2012). The impact of CO2

on economic growth in ASEAN 8. Procedia-Social and Behavioral

Sciences, 35, 389-397.

cause of rising CO pollution is very important as total emissions

rise sharply. Evidence from the findings shows that the main

actors or driving forces leading to a high level of environmental

damage are demographics, GDP per capita, and energy usage for

Malaysia and selected ASEAN+3 nations.

It can, therefore, be inferred that, with higher population size,

GDP per person, and energy consumption, the solution to

supporting a level of CO

will contribute to the rate of CO

Malaysia and Selected ASEAN+3 states, on the other hand, labor

productivity has no significant long-term effect on the rate of CO

emissions and reducing energy usage

emissions being regulated. In

Pesaran MH (2007) A simple panel unit root test in the presence of

cross-section dependence. J Appl Econ 22(2):265–312

Breitung, J., & Pesaran, M. H. (2008). Unit roots and cointegration in

panels. In The econometrics of panel data (pp. 279-322). Springer,

Berlin, Heidelberg.

pollution. Therefore, with the help of calculations, it can be

advocated but the use of further labor productivity during

industrial development would not cause deterioration of the

environment. Likewise, the workforce holding the real economy's

Breusch, T. S., & Pagan, A. R. (1980). The Lagrange multiplier test

and its applications to model specification in econometrics. The

review of economic studies, 47(1), 239-253.

0 Chandran, V. G. R., & Tang, C. F. (2013). The impacts of transport

energy consumption, foreign direct investment and income on CO2

emissions in ASEAN-5 economies. Renewable and Sustainable

Energy Reviews, 24, 445-453.

potential does not significantly contribute to a rise in CO

emission, which means that global economic growth will not

affect the rate of CO pollution. Therefore, the results suggest that

the main factors that can decrease CO emissions are energy

usage, level of public work and efficiency of labor. The actors of

expanded STIRPAT (population, economic growth, and energy

usage) in all Malaysia and selected ASEAN+3 nations support

long-term environmental damage.

11 Chen, Y. L., Analytis, J. G., Chu, J. H., Liu, Z. K., Mo, S. K., Qi, X.

L., ... & Zhang, S. C. (2009). Experimental realization of a three-

dimensional topological insulator, Bi2Te3. science, 325(5937), 178-

81.

Chudik, A., & Pesaran, M. H. (2015). Common correlated effects

estimation of heterogeneous dynamic panel data models with weakly

exogenous regressors. Journal of Econometrics, 188(2), 393-420.

Deltcheva, E., Chylinski, K., Sharma, C. M., Gonzales, K., Chao, Y.,

Pirzada, Z. A., ... & Charpentier, E. (2011). CRISPR RNA maturation

by trans-encoded small RNA and host factor RNase

III. Nature, 471(7340), 602.

These results provide a policymaker with useful information

to control both safe industrialization speed and rate of CO

emissions. Industrialization is fundamental element of

economic growth and cannot be limited, while the primary cause

for environmental damage is CO emissions. Lawmakers in

Malaysia and selected ASEAN+3 nations propose adding new

industrial units with environmentally-friendly technology that

consume low energy levels, leading to low levels of pollution.

Managing energy usage can be a significant contributor to

Dumitrescu, E. I., & Hurlin, C. (2012). Testing for Granger non-

causality in heterogeneous panels. Economic modelling, 29(4), 1450-

460.

5 Farhani, S., & Ozturk, I. (2015). Causal relationship between CO 2

emissions, real GDP, energy consumption, financial development,

trade openness, and urbanization in Tunisia. Environmental Science

and Pollution Research, 22(20), 15663-15676.

mitigating CO pollution, particularly for the industry sector in

Malaysia and selected ASEAN+3 nations, and this can be made

possible through effective policies with government support.

Such countries in Malaysia and Identified ASEAN+3 need to

implement more restrictive energy policies for short- and long-

term non-renewable sources of energy and should invest more in

research and development to incorporate environmentally

friendly sources of energy. Renewable and alternative energy

options that are an alternative to non-renewable energy (oil)

should be discussed being used on a massive scale. More notably,

renewable energy will eventually replace fossil fuel energy

because recent research and technological advances have resulted

16 Friedl, B., & Getzner, M. (2003). Determinants of CO2 emissions in

a small open economy. Ecological economics, 45(1), 133-148.

7 Go, F. M., & Govers, R. (1999, August). The Asian perspective: which

international conference destinations in Asia are the most

competitive?. In Journal of Convention

Exhibition

Management (Vol. 1, No. 4, pp. 37-50). Taylor & Francis Group.

8 Halicioglu, F. (2009). An econometric study of CO2 emissions, energy

consumption, income and foreign trade in Turkey. Energy

Policy, 37(3), 1156-1164.

Journal of Environmental Treatment Techniques

2020, Volume 8, Issue 1, Pages: 573-581

9 Hui, T. S., Rahman, S. A., & Labadin, J. (2012). Statistical modelling

of CO2 emissions in Malaysia and Thailand. International Journal on

Advanced Science, Engineering and Information Technology, 2(5),

41 Swamy, P. A. (1970). Efficient inference in a random coefficient

regression model. Econometrica: Journal of the Econometric Society,

311-323.

50-355.

42 Swamy, P. A. (1970). Efficient inference in a random coefficient

regression model. Econometrica: Journal of the Econometric Society,

311-323.

0 Iwata, H., Okada, K., & Samreth, S. (2010). Empirical study on the

environmental Kuznets curve for CO2 in France: the role of nuclear

energy. Energy Policy, 38(8), 4057-4063.

43 Thomou, T., Mori, M. A., Dreyfuss, J. M., Konishi, M., Sakaguchi,

1 Jalil, A., & Mahmud, S. F. (2009). Environment Kuznets curve for

M., Wolfrum, C., ... & Gorden, P. (2017). Adipose-derived

circulating miRNAs regulate gene expression in other

tissues. Nature, 542(7642), 450.

CO2 emissions:

cointegration analysis for China. Energy

policy, 37(12), 5167-5172.

Kaika, D., & Zervas, E. (2013). The Environmental Kuznets Curve

44 Wang, J. (2011). The end of the revolution: China and the limits of

modernity (Vol. 40, No. 5, pp. 631-633). Sage CA: Los Angeles, CA:

SAGE Publications.

45 Washington, W. M., Weatherly, J. W., Meehl, G. A., Semtner Jr, A.

J., Bettge, T. W., Craig, A. P., ... & Zhang, Y. (2000). Parallel climate

model (PCM) control and transient simulations. Climate

Dynamics, 16(10-11), 755-774.

46 Yusuf, S., Pfeffer, M. A., Swedberg, K., Granger, C. B., Held, P.,

McMurray, J. J., ... & CHARM Investigators and Committees.

(2003). Effects of candesartan in patients with chronic heart failure

and preserved left-ventricular ejection fraction: the CHARM-

Preserved Trial. The Lancet, 362(9386), 777-781.

(EKC) theory—Part A: Concept, causes and the CO2 emissions

case. Energy Policy, 62, 1392-1402.

Karki, S. K., Mann, M. D., & Salehfar, H. (2005). Energy and

environment in the ASEAN: challenges and opportunities. Energy

Policy, 33(4), 499-509.

Lise, W., & Van Montfort, K. (2007). Energy consumption and GDP

in Turkey: Is there

co‐integration relationship?. Energy

economics, 29(6), 1166-1178.

5 Liu, Z., Zhang, M., Bhandari, B., & Wang, Y. (2017). 3D printing:

Printing precision and application in food sector. Trends in Food

Science & Technology, 69, 83-94.

M. Hashem Pesaran, Yongcheol Shin & Ron P. Smith (1999) Pooled

Mean Group Estimation of Dynamic Heterogeneous Panels, Journal

of the American Statistical Association, 94:446, 621-

34, DOI: 10.1080/01621459.1999.10474156.

7 Masih, A. M., & Masih, R. (1996). Energy consumption, real income

and temporal causality: results from a multi-country study based on

cointegration and error-correction modelling techniques. Energy

economics, 18(3), 165-183.

Mericske‐Stern, R., Piotti, M., & Sirtes, G. (1996). 3‐D in vivo force

measurements on mandibular implants supporting overdentures. A

comparative study. Clinical oral implants research, 7(4), 387-396.

9 Oh, W., & Lee, K. (2004). Causal relationship between energy

consumption and GDP revisited: the case of Korea 1970–

999. Energy economics, 26(1), 51-59.

Okushima, S. (2016). Measuring energy poverty in Japan, 2004–

013. Energy policy, 98, 557-564.

Pedroni, P. (2004). Panel cointegration: asymptotic and finite sample

properties of pooled time series tests with an application to the PPP

hypothesis. Econometric theory, 20(3), 597-625.

Persyn, D.,

cointegration tests for panel data. The STATA journal, 8(2), 232-241.

Persyn, D., Westerlund, J. (2008). Error-correction–based

cointegration tests for panel data. The STATA journal, 8(2), 232-241.

Pesaran, M. H. (2004). General diagnostic tests for cross section

dependence in panels.

Roy, D. K., & Datta, B. (2017). Multivariate adaptive regression

spline ensembles for management of multilayered coastal

aquifers. Journal of Hydrologic Engineering, 22(9), 04017031.

Saboori, B., & Sulaiman, J. (2013). Environmental degradation,

economic growth and energy consumption: Evidence of the

environmental Kuznets curve in Malaysia. Energy Policy, 60, 892-

& Westerlund, J. (2008). Error-correction–based

05.

7 Soytas, U., & Sari, R. (2009). Energy consumption, economic growth,

and carbon emissions: challenges faced by an EU candidate

member. Ecological economics, 68(6), 1667-1675.

8 Soytas, U., & Sari, R. (2009). Energy consumption, economic growth,

and carbon emissions: challenges faced by an EU candidate

member. Ecological economics, 68(6), 1667-1675.

Soytas, U., Sari, R., & Ewing, B. T. (2007). Energy consumption,

income, and carbon emissions in the United States. Ecological

Economics, 62(3-4), 482-489.

Stocker, T. F., Qin, D., Plattner, G. K., Tignor, M., Allen, S. K.,

Boschung, J., ... & Midgley, P. M. (2013). Climate change 2013: The

physical science basis.