Globalisation and its effect on pollutio

In addition, this study uses Malaysia’s total trade with 10 TPP members as an indicator Malaysia.. The direct effects of industrial pro-duction, transportation and deforestation, which a

RESEARCH ARTICLE

Globalisation and its effect on pollution in Malaysia: the role

of Trans-Pacific Partnership (TPP) agreement

Received: 28 November 2016 / Accepted: 11 August 2017

# Springer-Verlag GmbH Germany 2017

Abstract The main objective of this study is to investigate

the i nfluence of the g lobal isat ion (Trans-Pacific

Partnership (TPP) agreement in particular) on air pollution

in Malaysia To achieve this goal, the Autoregressive

Distributed Lag (ARDL) model, Johansen cointegration

test and fully modified ordinary least square (FMOLS)

of pollution while GDP per capita and urbanisation serve

as its other determinants In addition, this study uses

Malaysia’s total trade with 10 TPP members as an indicator

Malaysia The outcome of this research shows that the

variables are cointegrated Additionally, GDP per capita,

urbanisation and trade between Malaysia and its 10 TPP

gen-eral Based on the outcome of this research, important

pol-icy implications are provided for the investigated country

Trans-Pacific Partnership agreement

Introduction

The rising inter-linkages among various economies across the world have multifaceted impacts on the socio-economic-political aspects of life Associated with these inter-linkages

is the rising volume of international trade, which has both positive and negative implications on the environmental qual-ity of the trading nations International trade can aid the pos-itive spread of environmentally friendly practices and technol-ogies from advanced to developing nations It can also de-crease pollution in emerging economies through, for instance, the importation of cleaner technologies or through the devel-opment of better environmental regulations and standards Nations can utilise trade relations to encourage their emitting neighbours to decrease pollution or join transnational

quality is considered a normal good, the demand for it will

hand, fossil fuels including coal, natural gas and oil are re-quired in the process of producing goods and services for international transactions The direct effects of industrial pro-duction, transportation and deforestation, which are associated with international trade, include emissions and environmental

that examine the effects of trade on the environment not to overlook the pollution haven hypothesis (PHH), which envis-ages that the removal of trade barriers makes dirty companies

to migrate to countries with loose environmental standards Developing nations that are poor usually function as pollution havens and thereby generate more emissions These poor na-tions trade more due to the availability of a neighbouring large

There are several trade agreements available in this era, and the main intention of these agreements is to increase interna-tional trade The growing realisation of the role of free trade

Responsible editor: Philippe Garrigues

* Sakiru Adebola Solarin

sasolarin@mmu.edu.my

Usama Al-mulali

usama.almulali@mmu.edu.my

Pritish Kumar Sahu

pritish.sahu@mmu.edu.my

1 Faculty of Business, Multimedia University,

75450 Melaka, Malaysia

Environ Sci Pollut Res

DOI 10.1007/s11356-017-9950-0

agreements in improving the overall economic performance of

the participant countries resulted in intensifying the trade

ne-gotiations across the globe since the late 80s Developing and

emerging economies are particularly interested in such

agree-ments as they expect to strengthen their market access,

eco-nomic growth, income level and living standards, to mention a

few

The Trans-Pacific Partnership (TPP) agreement is an

agreement signed on the fourth of February 2016 after

19 rounds of tough negotiations that took over 5 years

to achieve Regarded as the biggest trade agreement of

the twenty-first century, the TPP agreement is the

succes-sor of the Trans-Pacific Strategic Economic Partnership

Agreement, or TPSEP, which was signed by Brunei,

Chile, New Zealand and Singapore in 2005 The TPP

trade bloc comprises of 11 nations, including Australia,

Brunei, Canada, Chile, Japan, Malaysia, Mexico, New

account for over 14% of the global GDP and almost 7%

is one of the several Mega-Regional Trade Agreements

(MRTAs) that have emerged since the mid-1990s As a

deep and comprehensive trade agreement, the TPP covers

traditional barriers to trade in goods and services (e.g

tariffs, restrictions on the movement of professionals),

in-vestment activities and other trade-related areas Such

areas include formal restrictions on some trade and

invest-ment activities, burdensome and inconsistent regulations,

varying treatment of intellectual properties, differing

la-bour and environmental standards, issues specific to small

and medium-size enterprises and new challenges arising

from rapidly growing digital technologies (World Bank

2014)

This monumental trade deal raises some widespread

speculations as to how this would affect countries in the

Asian region, particularly the relatively smaller countries

Undoubtedly, it is evident that the free trade plays a

sig-nificant role to the economic growth and welfare of the

established a negative or insignificant impact of free trade

of TPP, there are limited empirical studies, mainly

be-cause of the confidentiality of the TPP (Cororaton and

the gains, losses and the economic welfare for several

participating and important non-participating countries in the region Mostly, these studies have estimated that Malaysia is likely to gain in economic fronts during the post-tariff elimination period The agreement is also ex-pected to generate a rise in the direct cost of medicine and

we are not aware of any empirical studies exclusively focusing on the effect of post-TPP tariff elimination on emissions

The objective of this paper is to examine the potential

of 1970–2014 We contribute to the existing literature on

Firstly, we use indices for TPP as additional determinants

proxies for international trade without considering the po-tential role of TPP on emissions Secondly, we incorpo-rate structural breaks in the estimation process, including the unit root testing and cointegration procedures Specifically, we introduce a residual augmented least squares (RALS) unit root test on the series involved in a trade-emission exercise The method is a powerful unit root testing method that provides for endogenously deter-mined structural breaks Unlike most of the existing line-arity tests, the method is still robust in the presence of nonlinearity The RALS procedure provides for any evi-dence of non-normality, including asymmetry and

We focus on Malaysia for two reasons Firstly, among the 11 TPP members, it is a typical example

of developing countries that are facing rising levels of emissions in spite of the various proactive actions of the governments to curb the menace Although Malaysia’s share in the global emissions is very low, the intensity levels of the country’s emissions are higher than the global average in the energy sector (Economic

Malaysia increased from 9.8 million tonnes in 1970 to

60 million tonnes in 1990 and further increased to 258 million tonnes in 2014 Among the efforts of the gov-ernment to reduce pollution was the introduction of the Environmental Quality Act of 1974 aimed at ensuring that the environment is clean, healthy and safe In order

to reinforce environmental regulations, the Act has been developed over the years In the transportation sector,

emis-sions in addition to encouraging a greater use of biofuels and energy efficient vehicles (Economic

diver-gence in opinion on the implications of the agreement

on the Asian economies, including Malaysia (Lee et al

1 The USA was initially a signatory to the agreement USA has formally

withdrawn from the agreement through presidential memorandum.

Therefore, it is virtually impossible that USA may ratify the agreement.

Environ Sci Pollut Res

present an important implication of the agreement on

Malaysia

Literature review

contributors to greenhouse gas emissions, which present a

major dilemma for the globe due to the increase in human

activities worldwide Therefore, a large number of

empir-ical studies examined the main factors that contributed to

the environmental pollution Numerous scholars had

al-ready made a detailed summary on the empirical studies

that examined the environmental pollution model during

Therefore, this research will provide a summary of the

investigated the main determinants of environmental

pol-lution Despite the different methods and countries the

previous studies had investigated, it is clear that real

2 0 1 6; Z h u e t a l 2 0 1 6; Do g a n a n d S e k e r 2 0 1 6;

degra-dation Regarding the other determinants, namely trade

openness and financial development, the results are not

uniform A number of studies found that trade openness

However, other scholars found that both variables

most of the scholars that examined the effect of renewable

energy consumption (clean sources of energy) on pollu-tion reached the same conclusion which indicated its sig-nificant effect on mitigating pollution levels (Al-mulali

re-cent considerable number of studies, the influence of globalisation on environmental degradation is rarely

re-sults of these studies are similar to the older research that investigated this relationship (Christmann and Taylor

of globalisation (export plus imports divided by total GDP) However, a more specific definition can provide

a better picture to understanding the relationship between globalisation and pollution

Methodology

Model and data The STRIPAT (which represents Stochastic Impacts by Regression on Population, Affluence and Technology) framework is adopted to examine the factors of environ-mental degradation or pollution in Malaysia According to the STRIPAT, the magnitude of the environmental quality

is shaped by affluence level or economic prosperity, de-mography and the level of technology in a country (Dietz

af-fluence or economic prosperity is reflected in the average propensity to consume (APC) As the average consump-tion in the economy rises, polluconsump-tion level also rises A

an economy’s production, it is usually expected that consumption rises when production rises (York et al

popula-tion and can represent the demographic changes of a country The process of urbanisation and the growth of cities are the result of increase in population and popula-tion density Technology signifies the other determinants

of environmental quality beyond affluence and

proxy of technology The transfusions of technological innovations are associated with the rapid rate of globali-sation Through technical and scientific seminars, media, internet and several other communication mechanisms, globalisation promotes knowledge transmission at a much greater pace compared to past experiences (Archibugi and

devel-oped countries to the developing ones is a common phe-nomenon of the contemporary era On the other hand, trade influences energy demand and environmental Environ Sci Pollut Res

O2

O2

O2

Environ Sci Pollut Res

O2

O2

O2

Environ Sci Pollut Res

Environ Sci Pollut Res

quality by transferring the pollution inclined technology

to nations where environmental regulations are feeble,

es-pecially in the underdeveloped economies Therefore, the

equation is specified as follows:

Malaysia’s total trade with the other 10 TPP countries (which

is Malaysia’s total exports into the other 10 TPP countries plus

globalisation in this paper In the subsequent equations, we

further use many proxies for globalisation including

Malaysia’s total trade with each TPP country as a share of

Malaysia’s GDP, Malaysia’s total trade with all trading

total trade with TPP countries plus the USA as a share of

in-cluded in the equation to capture the structural breaks

Although we provide for a maximum break of two periods,

the unit root tests show that there is one significant structural

single break in the bound test A similar paper that has used

emis-sions in several cases as they involve fuel combustion in the

residential, industrial, power generation and transportation

sectors, which increase greenhouse emissions (GHGs)

absorption by plants Natural processes, such as plant matter

decay, also cause pollutions Countries with fossil fuels

dominating their energy mix are likely to be experiencing

increases in emissions With energy inefficiencies and

prevalence of energy wastage, it is expected that economic

activities will lead to more emissions in the country

Previous studies that have real GDP in their pollution

Trade is, in some ways, a type of technology Moreover, it

is connected with human activities that cause emissions such

as transportation, industrial production and deforestation

developing countries has lured multinational corporations to

move their plants from high-income nations to low-income

countries and these companies pay poor salaries and do not

usually fulfil the environmental laws imposed by the

has been criticised on the basis that it encourages more pro-duction which negatively affects the environmental quality because of the poor production techniques (Ling et al

production bases to the developing countries because of slack

expected that the increase in trade will lead to more emissions

in the country Previous works that have added trade

Several emerging economies are undergoing economic transformation that will eventually cause physical expansion

of the urban centres The urban areas are usually energy-intensive with high propensity of economic activities, which are frequently driven by fossil fuels that cause environmental degradation The quick pace of urbanisation in recent decades will probably cause the snowballing of energy demand and pollution Therefore, it is expected that increase trade will lead

to more emissions in the country Previous works that have added urbanisation in the pollution equations include Solarin

Review of World Energy, while the data for GDP, exports and imports are collected from World Integrated Trade Solution (Wits) supplied by World Bank database for the period of

exports and imports were available for only the period of 1975–2014 The urban population ratio is collected from the world development indicators of the World Bank The data for population (which is used as a divisor in order to obtain emis-sion per capita, real GDP per capita and urban population ratio) is generated from world development indicators of the

Malaysia’s total trade with all TPP countries plus the USA

as a share of Malaysia’s GDP The other variables have been defined earlier The variables are reported in their original forms The mean statistics show that the trade between Malaysia and the TPP countries was almost 38% of the total trade of Malaysia in the period of 1970–2014 The

Jarque-Table 2 Descriptive analysis Series Mean Standard deviation Jarque-Bera

The parenthesis contains the probability values Environ Sci Pollut Res

Table 3 Two-break LM and RALS-LM unit root tests

Δ ln Australia t −5.844 a

Δ ln Brunei t −8.844 a

Δ ln Canada t −8.941 a

[0] −10.331 a

Δ ln Chile t −9.812 a

[0] −10.986 a

Δ ln Japan t −6.706 a

Δ ln Mexico t −4.978 a

Δ ln Singapore t −5.065 a

Δ ln Vietnam t −6.989 a

Δ ln World t −5.528 a

t

−6.298 [1] −12.122 a

Due to the fact that the LM test and RALS-LM test similarly share the same process to search for the break points and the relevant optimal lags, we only report one time to conserve space The optimal number of lagged first-differenced term is reported in the parenthesis TB is the structural break point(s) The critical values are based on Akaike Information Criterion (AIC) The critical values of the LM test for two breaks are −4.689, −4.183 and −3.921 at the 1, 5 and 10% levels, respectively The critical values of the LM test for one break are −4.199, −3.671 and −3.403 at the 1, 5 and 10% levels, respectively All the critical values are computed, using the codes provided in https://www.dropbox.com/sh/dnjpjqmmgfi4otu/

a

1% significance level

b

5% significance level

c

10% significance level

Environ Sci Pollut Res

Bera statistics suggest that all the variables follow normal

distribution

Unit root tests

The RALS-LM unit root test is based on the following

regres-sion:

Rt; DT*

Rt

transformation is needed to remove the dependency of the test

(contains the information on non-normal errors in a bid to

Table 4 Bounds test

Normal

lnEMI t = f(lnRGDP t , lnURB t , lnTPP t , DUM t ) 9.916 c (1,4,3,0) 0.741

[2]

0.136 [1]

0.689 [1] lnEMI t = f(lnRGDP t , lnURB t , lnAUSTRALIA t , DUM t ) 5.751 b (1,0,4,1) 0.722

[1]

0.579 [1]

0.471 [2] lnEMI t = f(lnRGDP t , lnURB t , lnBRUNEI t , DUM t ) 4.302 a (1,0,4,3) 0.151

[1]

0.580 [1]

0.621 [2] lnEMI t = f(lnRGDP t , lnURB t , lnCANADA t , DUM t ) 5.829 b (1,0,3,2) 0.398

[1]

0.265 [1]

0.413 [2] lnEMI t = f(lnRGDP t , lnURB t , lnCHILE t , DUM t ) 4.369 a (1,3,3,0) 0.115 [1] 0.689

[1]

0.382 [2] lnEMI t = f(lnRGDP t , lnURB t , lnJAPAN t , DUM t ) 7.936 c (1,0,2,2) 0.513

[1]

0.716 [1]

0.607 [2] lnEMI t = f(lnRGDP t , lnURB t , lnMEXICO t , DUM t ) 4.237 a (1,0,4,0) 0.813

[1]

0.301[1] 0.537

[2] lnEMI t = f(lnRGDP t , lnURB t , lnNEWZEALAND t , DUM t ) 6.369 c (1,0,4,1) 0.799

[1]

0.505 [1]

0.197 [2] lnEMI t = f(lnRGDP t , lnURB t , lnPERU t , DUM t ) 5.261 b (1,0,4,0) 0.136

[1]

0.638 [1]

0.741 [2] lnEMI t = f(lnRGDP t , lnURB t , lnSINGAPORE t , DUM t ) 8.135 c (1,0,3,0) 0.733

[1]

0.313 [1]

0.292 [2] lnEMI t = f(lnRGDP t , lnURB t , lnVIETNAM t , DUM t ) 4.512 a (3,0,0,0) 0.112 [2] 0.163

[2]

0.706 [2] lnEMI t = f(lnRGDP t , lnURB t , lnWORLD t , DUM t ) 7.112 c (1,0,3,1) 0.359

[1]

0.501 [1]

0.254 [2] lnEMI t = f(lnRGDP t , lnURB t , lnUTPP t , DUM t ) 8.111 c (1,0,3,0) 0.348

[1]

0.387 [1]

0.354 [2]

[1]

0.136 [1]

0.112 [2]

For the four-variable models, the critical values (for lower and upper bounds) are (5.150, 6.280), (3.822, 4.714) and (3.226, 4.054), at 1, 5 and 10%, respectively For the five-variable model, the critical values (for lower and upper bounds) are (4.628, 5.865), (3.470, 4.470) and (2.950, 3.862), at 1, 5 and 10%, respectively The brackets show the order of diagnostic tests The specifications include unrestricted intercept and restricted trend The breaks included in the model is dummy for 1994

a

1% significance level

b

5% significance level

c

10% significance level

Environ Sci Pollut Res

further improve the strength of the LM statistic) is introduced

equations to ensure that there are no autocorrelations in the

models

ARDL bound test

To employ the Autoregressive Distributed Lag (ARDL)

unrestricted error correction model (UECM) is estimated:

ΔlnEMI t ¼ α 1 þ ∑k

i¼1 α 2 ΔlnRGDP t−i þ ∑k

i¼0 α 3 ΔlnURB t−i þ ∑k

i¼0 α 4 ΔlnTPP t−i

α 5 EMIt−1þ α 6 RGDPt−1þ α 7 URBt−1þ α 8 TPPt−1þ α 9 Dtþ υ t

ð3Þ

T h e n u l l h y p o t h e s i s o f n o c o i n t e g r a t i o n

relationship between the variables, we estimate the short run

model as specified as follows:

ΔlnEMI t ¼ α 1 þ ∑k

i¼1 α 2 ΔlnRGDP t−i þ ∑k

i¼0 α 3 ΔlnURB t−i

þ ∑k

i¼0 α 4 ΔlnTPP t−i þ α 5 D t þ α 6 ECTt−1þ υ t

ð4Þ

residual generated from the estimation of the cointegration

equation in the long run models Finally, the stability of the

models is analysed through the cumulative sum (CUSUM)

and cumulative sum of squares (CUSUMSQ) tests

Results

The empirical analyses commence by testing the unit root

study and, for comparison, we also report the test

is utilised to determine the optimal lag In their original

forms, nonstationarity is supported for all the variables

When the variables are entered in their first differences,

stationarity is supported for all the variables at 10% level

or better It is observed that 27% of the structural breaks occurred in the latter part of 1990s, which is the period associated with the Asian financial crisis The crisis, which started because of speculative attacks on national currency of Thailand (Baht), spread to other neighbouring countries and affected not only the financial sector but also real sector in Malaysia Domestic-oriented sectors such as construction and services sectors were also ad-versely affected Malaysia experienced the biggest plunge

in the region as stock market capitalization decreased by about 76% Therefore, several sectors, including the con-struction and services industries, were harshly affected by

After observing the integration properties of the series,

we proceed with the ARDL test to examine potential long

findings of different set of equations In the first model,

emis-sion (as the dependent variable), real GDP, urban popula-tion ratio and real trade per capita with TPP members The evidence suggests that there is cointegration as the

F statistics (9.916) is greater than the upper critical value (6.280) at 1% significance level We further examine the cointegration while using the real trade per capita in each TPP members as proxies for trade openness It is observed that we cannot accept the null of no cointegration when the dataset of each TPP country is utilised as proxy for

t r a d e o p e n n e s s F u r t h e r m o r e , w e a l s o t e s t f o r cointegration, when the total trade for all trading countries and TPP members plus the USA are used The F statistics

in the equation involving all trading countries (7.112) and the F statistics in the equation involving TPP countries plus the USA (8.111) are bigger than the upper critical value at (6.280) 1% significance level Lastly, we examine the possibility of cointegration in an equation involving

GDP square, urban population ratio and real trade per capita with TPP members The evidence suggests that there is cointegration as the F statistics (4.844) is greater than the upper critical value (4.470) at 5% significance level The diagnostic tests indicate that there is no prob-lem of serial correlation, heteroscedasticity and non-normality

coefficients and the focus is on the equations with evidence for cointegration In the first model, we report the findings of

vari-able), real GDP, urban population ratio and real trade per capita with TPP members We observe that real GDP, urban population ratio and real trade per capita with TPP members

significance level or better In model 2 to model 11, we replace real trade per capita with TPP members with real trade per

2 It has been demonstrated that the asymptotic distribution of τ × RALS is

given as follows:τ*RALS−LM→ρ~τ*LMþpffiffiffiffiffiffiffiffiffiffiffiffiffi1−ρ2Z

rel-ative ratio of the variances of two error terms

Environ Sci Pollut Res