Renewable Energy 189 (2022) 454e466

Contents lists available at ScienceDirect

Renewable Energy
journal homepage: www.elsevier.com/locate/renene

The impact of renewable energy, internet use and foreign direct

investment on carbon dioxide emissions: A method of moments
quantile analysis
Ashar Awan a, b, Kashif Raza Abbasi c, d, *, Soumen Rej e, f, Arunava Bandyopadhyay f,
Kangjuan Lv g, **
a
NisanTasi University, Graduate School, Turkey
b
University of Azad Jammu and Kashmir, Pakistan
c
School of Economics, Shanghai University, No. 99, Shangda Road, Baoshan Campus, Baoshan District, Shanghai, 200444, China
d
Department of Business Administration, Faculty of Management Sciences, ILMA University, Karachi, Pakistan
e
GITAM Institute of Management, Gandhi Institute of Technology and Management, Visakhapatnam, 530045, India
f
Vinod Gupta School of Management, Indian Institute of Technology Kharagpur, West Bengal, Kharagpur, India
g
SILC Business School, Shanghai University, 20 Chengzhong Road, Jiading Dist, 201899, Shanghai, People's Republic of China

a r t i c l e i n f o a b s t r a c t

Article history: The current era of industrial development and innovation is revolutionized using the internet, robotics,
Received 2 November 2021 and artificial intelligence. Nevertheless, the appraisal of global economic progress shows increasing
Received in revised form trends, there are environmental degradation issues associated with this improvement. The role of
28 January 2022
renewable energy, urbanization and foreign direct investment received a lot of attention in the literature
Accepted 3 March 2022
Available online 9 March 2022
on environmental issues, however, the simultaneity with information communication technology is
missing. Therefore, the present study used data from 10 emerging countries during 1996e2015 and
applied the novel Method of Moments Quantile Regression to analyze the nexus among the variables.
Keywords:
Renewable energy
Further, we applied the second-generation unit root test, and Driscoll Kraay standard errors to reach
Foreign direct investment robust results. The findings revealed an inverted U-shape relationship between economic growth and
The environmental kuznets curve environmental degradation; thus, the validity of the Environmental Kuznets Curve is revealed. Moreover,
Urbanization foreign direct investment is significant and positive at 0.05th-0.50th quantiles, however, it becomes
Information communication technology insignificant at higher quantile levels. Urbanization enhances while renewable energy mitigates carbon
dioxide emissions at all quantile levels. Information communication technology proxied by internet
usage reduces environmental degradation significantly at 0.25th-0.95th quantile levels. Results of the
study suggest insights for the policymakers to mitigate carbon dioxide emissions through encouraging
renewable energy and internet use.
© 2022 Elsevier Ltd. All rights reserved.

1. Introduction caused such environmental degradation. Among various green-

house gases, carbon dioxide (CO2) emissions is the major contrib-
The theoretical environmental concerns have become a real- utor and is now the target of different global environmental
world issue for recent and future generations. The massive prog- agendas. Earlier literature has discussed various factors that impact
ress, increase in production and consumption of industrial output CO2 emissions, including renewable energy use, economic growth,
financial development, urbanization [1e5]. Perhaps renewable is
the most plausible solution to curb CO2 emissions. Nevertheless,
* Corresponding author. School of Economics, Shanghai University, No. 99, the impact of energy consumption on the environment is damaging
Shangda Road, Baoshan Campus, Baoshan District, Shanghai, 200444, China. and reducing energy consumption could be a disaster for economic
** Corresponding author. SILC Business School, Shanghai University, 20 Chengz- growth. However, renewable energy is an environmental-friendly
hong Road, Jiading Dist, 201899, Shanghai, People's Republic of China.
energy source. Renewable energy is promising in a sense that it is
E-mail addresses: asharawan786@hotmail.com (A. Awan), kashifabbasi@shu.
edu.cn (K.R. Abbasi), soumen.rej@iitkgp.ac.in (S. Rej), arunava.b@iitkgp.ac.in supportive to the environment, low waste production is involved in
(A. Bandyopadhyay), lvkangjuan@shu.edu.cn (K. Lv). its generation, and it ensures reliable energy supply. Previous

https://doi.org/10.1016/j.renene.2022.03.017
0960-1481/© 2022 Elsevier Ltd. All rights reserved.
A. Awan, K.R. Abbasi, S. Rej et al. Renewable Energy 189 (2022) 454e466

literature argued that achieving environmental sustainability is heterogeneity. This technique allows for diverse environmental-
impossible without promoting renewable energy [6e9]. Earlier emissions linkages at multiple conditional quantiles distributions,
literature reported Renewable energy reduces CO2 in the EKC not captured by traditional mean regressionsdpanel estimate
framework [10,11]. methods, including the novel "Method of Moments Quantile
On the other hand, one of the primary ingredients of economic Regression" (MMQR). In addition, the fully modified ordinary least
growth is the Information and communication technologies (ICT) square (FMOLS) is utilized with the dynamic ordinary least square
that accelerate economic growth [12,13]. The Internet, mobile, (DOLS), which accounts for the endogeneity concerns in cross-
telephone, computer systems, and associated applications, collec- sectional data, and the Dumitrescu Hurlin panel causality are
tively known as ICTs, have become the primary drivers of societal other unique features of the analysis. The estimation findings
transformation, growth, and invention. On the other hand, the provide policy implications based on the nexus between FDI, ICT,
broad usage of ICT contributes to increased production efficiency. In renewable energy, and environmental policy for emerging coun-
addition, ICT has contributed to significant systemic trans- tries. In addition, Figs. 1e6 shows the recent trends of FDI, ICT,
formations in the international economy over the last few decades economic growth, urbanization, environmental degradation, and
by building smart cities, health systems and transportation systems renewable energy in 10 emerging countries. China and India show a
[14] and through encouraging global supply chains and boosting solid upward trend in environmental degradation and economic
economic development [15]. growth. In the sample period, Turkey, Hungary, and India attracted
While ICT is regarded as a stimulus to economic growth, and considerable FDI. The process of urbanization in China remained
subsequently, its environmental impacts are keenly investigated. remarkable during this period. The share of renewable energy in
ICT has been recognized as a vital factor to achieve a low carbon total energy consumption increased in Hungary, Romania, and
economy [15]. Earlier research argued that ICT might have a good or Bulgaria. Internet usage shows an increasing trend in all countries.
detrimental impact on environmental stability. ICT, for instance, The research contributes to the environmental literature in a
improves environmental protection by lowering CO2 emissions variety of ways. To our best knowledge, this is the first research of
[16]. Even though [17] claimed that Internet commerce reduces CO2 its kind that rigorously assesses the role of renewable energy, FDI,
emissions in general. The impact of the Internet on CO2 emissions, ICTs, and urbanization in CO2 emissions for emerging countries. For
on the other hand, varies significantly between industrialized and emerging economies, the nexus between ICT, urbanization, and
developing nations. renewable energy with the EKC hypothesis has not received due
Similarly, ICTs aid in the mitigation of atmospheric greenhouse attention, therefore, extending the previous literature with more
gases (GHG) in two directions: by boosting the productivity of advanced econometric techniques is required for a detailed
existing procedures and by causing alternative consequences that analysis.
occur in more energy-efficient output and utilization processes Following the introduction, the body of the article is organized
[18]. In general, ICTs are one of the causes that contributes to as follows: section 2 “literature review,” section 3 “data model and
accelerated CO2 emissions through the manufacture of ICT gear and econometric approaches” section 4 “result and discussions” last
devices, energy use, and electronic material recycling. Moreover, section is conclusions and policy implications.
ICT use enhances energy demand, as documented by Refs. [19,20].
Furthermore, urbanization may have a positive or negative role 2. Literature review
in CO2 emission [21], as discussed by urbanization theories,
including modernization theory and environmental transition Earlier literature has linked CO2 emissions to a variety of factors
theory. Earlier literature shows that with ICT, urbanization becomes including GDP, financial development, nonrenewable and renew-
more efficient and fast [22]. With the rise of urbanization, demand able energy use [1e5]. Before analyzing the dynamics of carbon
for energy increases due to electronic devices and automation dioxide emissions, renewable energy consumption with an
machines usage [23], therefore, energy, urbanization and internet emphasis on internet usage FDI, economic growth and urbaniza-
usage are linked [20,24]. However, their environmental conse- tion, it is vital to discuss the scientific contribution related to this
quences require a careful analysis [25]. theme in the previous literature. The extant literature related to our
In addition, the influx of FDI may result in increased CO2 emis- study can be classified into three strands.
sions in the host nation, particularly in states that desperately seek The first strand investigated the relationship between economic
economic prosperity and a loosening of environmental restrictions development and environmental pollution. This strand mainly
to lure foreign capital. On the other hand, FDI inflows may result in follows [29] to examine the environmental impact of economic
beneficial knowledge transfer and headwinds, efficient energy use, development. This environment and development nexus was
and lower CO2 emissions in host nations. Thus, understanding the named the Environmental Kuznets Curve (EKC) after the famous
intricate connections between FDI inflows and CO2 emissions is study by Simon Kuznets [30]. Later a plethora of research followed
critical for providing better policy guidelines to these issues. the EKC hypothesis to study the environment and income growth
Previous literature shows that the link between renewable en- relationship [31e35]. Interestingly, this strand has two contrasting
ergy, ICTs and CO2 emissions is conflated and needs more consid- conclusions; on one hand, studies are promoting the idea of the EKC
eration. Hence, this study aims to further investigate the impact of hypothesis [36,37]. However, on the other hand, there are findings
ICTs on CO2 emissions in emerging economies by integrating FDI, by others that support the idea of invalidity of the EKC hypothesis
urbanization, GDP, GDP2, and renewable energy into the EKC [27,38e47].
framework. The global percentage of CO2 emissions emitted by The second group of the literature analyzed the linkage between
emerging economies increased from 48% in 1980 to 61% in 2011 renewable energy and economic growth. Some studies found
[26]. In addition, previous literature utilized the STIRPAT model to bidirectional causality among renewable energy and economic
analyze ICT, urbanization and CO2 nexus [25]. However, our study development [48,49]. In a recent study, renewable energy con-
used the EKC framework. Thus our study will contribute to the sumption and economic development are reported to have a bi-
debate initiated by Ref. [27] on EKC vs STIRPAT efficiency. directional causality for a sample of 58 countries [50]. In contrast,
This research adds to the thin body of literature on the nexus some other studies reported no causal association concerning
above using the MMQR [28]. Combining this technique with fixed economic growth and renewable energy, for instance Ref. [51]
effects makes it simpler to gain a practical grasp of the association's found no causality in a sample of European countries. One can find
455
A. Awan, K.R. Abbasi, S. Rej et al. Renewable Energy 189 (2022) 454e466

Fig. 1. Trends in carbon dioxide emission.

in previous literature, that unidirectional causality is reported also addition, ecological modernization theory also supports this idea of
[52]. improvement in the environment at later stages of development
Besides, economic growth and renewable energy, the EKC due to efficient technologies [53].
literature also employed other aspects of economies. For instance, The current study also included FDI in the model as an inde-
FDI, urbanization and internet usage were found by previous pendent variable. FDI has been a central focus of studies targeting
literature as explanatory variables of environmental pollution. the environment and economic development nexus. Foreign direct
Literature on urbanization and environmental degradation can be investment in many of the developing countries is attracted by
broadly viewed into three distinct groups. The first reported a weak environmental regulations. Particularly, countries that are
positive effect of urbanization on environmental degradation [53]. rich in natural resources with weak environmental regulatory laws
As more population shifts from rural to urban area, demand for are attracting FDI [56]. Literature on the nexus of FDI, Economic
energy increases due to the rise in vehicles and residential build- growth and environment can be classified into two major groups.
ings. This results in the deterioration of the environment. Envi- First; studies that found a positive effect of FDI on the environment
ronmental transition theory justifies such linkage between called the “pollution halo effect” [57]. And the second group that
environment and urbanization. In a recent study, urbanization is found FDI mitigating environment degradation, called the “pollu-
found to degrade the environment in a sample of 99 countries [54]. tion heaven effect” [58]. In a recent study [59], analyzed the role of
However, the notion of agglomeration disagrees with the FDI and ICT on CO2 for Pakistan during 1990Q4-2018Q1 using the
increasing effect of urbanization process on the environment and NARDL model. The findings of this study suggested that local pro-
argues for a mitigating effect. As the income level rises and ur- duction of ICT equipment improves the environment as compared
banization increases in an economy, fewer resources are required to importing them. Furthermore, their study endorsed the validity
for transportation and accommodation. Therefore, due to urban of the EKC for Pakistan. The study found a positive effect of FDI on
agglomeration, the consumption of energy reduces [55,95,96]. In environmental degradation in the long run.

456
A. Awan, K.R. Abbasi, S. Rej et al. Renewable Energy 189 (2022) 454e466

Fig. 2. Trends of gross domestic product.

Fig. 3. Trends in foreign direct investment.

457
A. Awan, K.R. Abbasi, S. Rej et al. Renewable Energy 189 (2022) 454e466

Fig. 4. Trends in urbanization.

Among many technologies, the internet is perhaps the most the relationship between renewable energy, FDI, internet usage and
effective technology that reshapes societies, particularly in urban urbanization with the EKC in emerging economies.
areas. One of the earlies studies on the role of information tech-
nology was [60]. This study offered a new perspective in environ- 3. Data and methodology
mental analysis and introduced information technology as a new
determinant. Later [61], offered a detailed framework for further For empirical investigation, the current study considers 20 years
investigations about the impact of ICT in sustainability. In the same of annual time series data spanning from 1996 to 2015 for con-
vein [62], analyzed the role of e-commerce on the environment and structing a balanced panel of 10 emerging economies (Argentina,
found a positive impact of former on later. Similarly [63], found that Bulgaria, China, Hungary, India, Mexico, Poland, Romania, Russian
reading online newspapers reduces environmental degradation. Federation, and Turkey). We have included these 10 emerging
Later studies also found environmental benefits of ICT, for instance economies in our present study as they have surpassed the other
Refs. [64e66] through energy efficiency [67], by reducing energy emerging economies CO2 emissions within the sample period.
demand [68,69] energy saving by using GPS in the transport sector Moreover, the choice of sample selection is purely based on data
[70], through enhancing online retailing. However, the extant availability and for the purpose of a balanced panel data. The
literature is conflated about the role of internet usage and other multivariate framework of the present study utilizes CO2 emissions
forms of ICT in environmental degradation. On the one hand, there to capture environmental degradation measured in metric tons per
are empirical findings that supports the idea of a harmful effect of capita, real GDP as a proxy of economic well-being measuring in per
ICT on environmental degradation [63,64,68]. While, other studies capita, consumption of renewable energy measured as percentage
argue that an increase in internet users, increase demand for more of total consumption, foreign direct investment captured through
electricity and electricity production puts pressure on the envi- net inflows in terms of percentage of GDP, urbanization (percentage
ronment [71,72]. of the total population) and individuals using the internet as a
Thus from the literature review, three gaps are visible; firstly proxy of internet penetration measuring in percentage of the total
there is no study on the role of ICT, FDI and urbanization in the EKC population. The data source for all the series involved in the study
framework, Secondly, in the case of emerging economies, there is are retrieved from the World Bank. Each of the variables has gone
no study on the emission-growth nexus, particularly concerning through normal logarithmic change to actuate stationarity prop-
ICT and urbanization. Thirdly, with a focus on cross-sectional erties and stay away from the issues of distributional properties.
dependence, the application of advanced models that analyze the The comprehensive description of variables is enlisted in Table 1.
heterogeneous effects of independent variables on various levels of Table 2 uncovers that all the variables follow left-skewed dis-
CO2 emissions is a key gap in the extant literature on the EKC tribution, except renewable energy consumption and FDI following
hypothesis. right-skewed distribution. All the variables except renewable en-
From the detailed review of literature presented above, it is ergy consumption are demonstrated not to follow a normal dis-
evident that there is vitally important to guide policymakers about tribution as evidenced from the Jarque-Bera test. These results
458
A. Awan, K.R. Abbasi, S. Rej et al. Renewable Energy 189 (2022) 454e466

Fig. 5. Trend in Internet users.

support the idea of using nonlinear models that can capture the knowledge in the event of creating non-crossing estimates in
heterogenous relationship among the factors of dependent quantile regression. Although fixed effects cannot account for
variable. heterogeneity, yet the application of MMQR covers this issue due to
its ability to render heterogeneous estimates across the entire
distribution. Moreover, the heterogeneous values of coefficients
3.1. Panel estimation techniques explicitly claim that MMQR deals with the issue of heterogeneity.
Conducting error analysis (using Mean Square Error) Machado and
Concocted by Machado and Silva [28], the present study Silva [28] showed through simulation results that, MMQR method
employed MMQR model to investigate the heterogeneous and provides more robust estimates as compared to conventional
distributional impacts across quantiles among CO2, real GDP models. Other methods of panel regression such as FMOLS and
growth, renewable energy, FDI, urbanization and internet usage. DOLS have advantage of dealing with serious correlation and
While conventional panel quantile regression techniques [73,74] endogeneity, however, they also lack from estimation based on
can provide reliable estimates in the presence of outliers and condition in the data. Therefore, the method is preferred because it
appropriate in a situation when conditional means of two variables deals with heterogeneity and endogeneity by considering asym-
have a weak association, however, they are unable to detect the metric and non-linear association among CO2 emissions and its
possible presence of unnoticed heterogeneity across the panel cross determinants. The conditional quantile estimates Qy(trX) of the
sections. Authorizing the individual fixed effect, MMQR method model of location-scale variant can be expressed by the following
unavoidably examines the conditional heterogeneous covariance equation:
impact of CO2 emission's determinants to effect the whole distri-
bution. In addition, rather shifting means like Koenker [73], Canay yit ¼ ai þ Xit ‘b þ ðdi þ Zit ‘gÞUit (1)
[74], it captures the covariance effect in overall distribution. The
advantage of utilizing this method is that it addresses the Where, the probability P fdi þZ‘it g > 0g ¼ 1 , (a, b’, d, g’)’ are the
conceivable presence of endogenous properties in the explanatory parameters to be estimated. The individual i fixed effects are
variables. This technique is likewise appropriate in such situations designated as (ai, di), i ¼ 1, ….., n and k vector of known elements of
where individual effects submerge the panel data model. Moreover, X is signified by Z, which are differentiable conversions with
the MMQR model also produces reliable estimates in the case of the component l mentioned below:
non-linear model. As compared to the nonlinear models such as
“Nonlinear Autoregressive Distributed Lag” NARDL, the advantage
Zl ¼ Zl ðXÞ l ¼ 1; :::; k (2)
of using MMQR is that it defines the threshold through data driven
process not exogenously [75]. Moreover, MMQR allows for asym- Xit is independently and identically distributed for any fixed i
metries based on location. MMQR can also provide huge bits of and also across time t. Uit is also independently and identically
459
A. Awan, K.R. Abbasi, S. Rej et al. Renewable Energy 189 (2022) 454e466

Fig. 6. Trends in renewable energy.

Table 1
Description of variables.

Variable Description of variables Measurement Source

LnFDI Foreign direct investment Net inflows (% of GDP) WDI

LnUrban Urban population Urbanization (% of total population WDI
LnCO2e Environmental pollution Carbon emission (mto in per capita form) WDI
LnGDP Economic Development GDP in US dollars (constant-2010) WDI
LnREN Renewable Energy As percentage in total energy cons. WDI
LnInternet Internet penetration Individuals using internet (% of population) WDI

Note: WDI¼The World Bank.

Table 2
Descriptive statistics.

Variables lnCO2e lnGDP lnGDPsq lnren lnfdi lninternet lnurban

Mean 1.492 8.756 77.224 2.467 0.986 2.328 4.105

Std. Dev. 0.598 0.752 12.361 0.708 0.884 1.798 0.316
Min 0.235 6.568 43.138 1.172 1.88 4.337 3.289
Max 2.454 9.606 92.271 3.985 3.921 4.326 4.516
Skew. 0.993 1.468 1.327 0.17 0.072 1.188 1.189
Kurt. 4.195 4.181 3.799 2.677 4.368 3.955 3.584
JB 44.3 83.2 63.05 1.6 15.4 54.1 49.5
Probability 0.000 0.000 0.000 0.423 0.000 0.000 0.000
Observations 200 200 200 200 200 200 200

distributed across individuals i through time t and are orthogonal to internet. Qy ðtjXÞ postulates that the structural quantiles are
Xit and are standardized to accomplish the moment conditions. distributed to the dependent variable yit(CO2 emissions) relying on
Equation (1) derives the following: the distribution (location) of exogenous variables Xit. Individual (i)
quantile (t) fixed effects is demonstrated by the scalar coefficient
Qy ðtjXÞ ¼ ðai þ di ðtÞÞ þ X‘it b þ Z‘it gqðtÞ (3) denoted as ai ðtÞ ¼ ai þ di qðtÞ. Intercept shift does not represent
the individual effect in contrast to the typical fixed least-squares
Where, x'it signifies a vector of independent variables that effects. Such parameters are time-invariant with heterogeneous
augmented GDP, GDP2, renewable energy, FDI, urbanization and effects that are suitable to diverge along the conditional

460
A. Awan, K.R. Abbasi, S. Rej et al. Renewable Energy 189 (2022) 454e466

distributional quantiles of the endogenous variable. The t-the Table 4

sample quantile represented by q(t) can be assessed by addressing Cointegration test Westerlund (2007).

the resulting optimization problem written in equation (4): Statistic Value Z-value P-value

Gt 3.222 1.261 0.000

XX  Ga 4.239 2.508 0.994
minq rt ðRit  ðdi þ Z‘it gÞq (4) Pt 3.24 3.277 0.001
i t
Pa 3.493 1.134 0.872

Where rt ðAÞ ¼ ðt 1ÞAIfA  0gTAIfA > 0g indicates the check Note: (H0): No cointegration; (Ha): at least one of the cross-sectional units has
cointegration (Gt and Ga) or Cointegration for the panel as a whole (Pt and Pa).
function.

4. Empirical results emission by 0.037%. All the coefficients are significant at 1% level of
significance. The robustness of the results is reviewed by excluding
4.1. Testing unit root the ICT variable and presented in Table 6. It can be observed that
coefficient signs and size are almost similar to the results presented
The stationary properties of the variables are evaluated using in Table 5. Moreover, the results are also statistically significant at
second-generation unit root tests to avert the problem of spurious 1% level.
regression in subsequent estimations. The present study applied
panel unit root test by Pesaran [76] and outcome is presented in 4.4. Method of Moments Quantile Regression
Table 3. The Pesaran [76] panel unit root test (CIPS) is conducted in
levels and first difference. Results show that only FDI and ICT are The results for Method of Moments Quantile Regression
non-stationary at levels however, all the series are stationary at the (MMQREG) are presented in Table 7. The coefficient of renewable
first difference, thereby signifying that all the variables are either energy consumption (REC) is negative and statistically significant at
I(0) or I(1). a 1% level of significance in all quantiles. The coefficient value varies
from 0.264 in 5th quantile to 0.2529 in 95th quantile, thereby
4.2. Panel cointegration test suggesting that a 1% increase in renewable energy consumption
aids in reducing the CO2 emission (CO2e) by approximately 26% in
To identify the long-term cointegrating relationship between emerging economies. It can be observed that the effect of economic
the variables, subsequently, the bootstrapped panel cointegration growth on CO2e is positive and highly significant in all the quan-
test of Westerlund [77] is employed. The results are presented in tiles. Moreover, the GDP square term (GDP2) is negative and highly
Table 4. The bootstrapped panel cointegration test of Westerlund significant in all quantiles. Juxtaposing the results, the coefficients
[77] shows evidence of long term cointegrating relationship. The of GDP and its square are positive, and negative respectively,
null hypothesis of no cointegrating relationship is rejected at a 1% thereby suggesting the existence of “Inverted U-shape EKC” for the
level of significance for all the tests. The robustness of the cointe- panel of 10 countries for the study period of 1996e2015. The co-
gration relationship is confirmed by the Kao [78] test for panel efficient of FDI is positive and statistically significant at a 10% level
cointegration. It can be observed that at 1% level of significance the for the lower quantiles (5th, 25th, and 50th), thereby highlighting
alternative hypothesis is accepted. The findings reveal that the that a 1% increase in FDI increases CO2e by 0.0125%e0.02% for
variables have a long-term association, verifying the validity and emerging economies. The coefficient of urbanization is significant
consistency of empirical findings. and positively correlated with environmental degradation in all
quantiles, thereby highlighting that a 1% increase in urbanization
4.3. Panel regression estimates can deteriorate the environment by 0.81e0.83% across all the
quantiles of carbon emission. Further, the coefficients are following
The results of panel regression models are presented in Table 5 a rising trajectory which indicates that higher CO2e quantiles are
using FMOLS and DOLS. Furthermore, to reach robust results, more affected due to urbanization-driven environmental degra-
subsequently, we estimated FE-OLS model with robust Driscoll- dation. However, the coefficients of ICT are negative and significant
Kraay standard errors are presented in Table 6. It can be observed for all quantiles except for the 5th quantile and reveal that a 1%
that coefficients of GDP and GDPsq are positive, and negative increase in ICT reduces carbon emission by 0.0125%e0.0254%. The
respectively, thereby suggesting the existence of “Inverted U-shape coefficient size gradually becomes more negative from lower to
EKC” for the panel of 10 countries for the study period of higher quantiles. Fig. 7 also shows the graphical representation of
1996e2015. Further, a 1% increase in FDI and Urbanization con- coefficients for various factors at different quantile levels.
tributes positively to the carbon emission by 0.008% and 1.244%,
however, consumption of renewable energy reduces carbon emis-
4.5. Discussion of the outcome
sion by 0.29%. However, ICT shows evidence of a negative rela-
tionship with CO2e, whereby 1% increase in ICT reduces carbon
The findings of the method of moment and fixed effect regres-
sion, both indicate that internet usage has a statistically significant
Table 3 negative impact on CO2e, thereby indicating that internet usage
Unit root test (Pesaran CADF). contributes in the improvement of air quality for emerging econ-
omies. This findings is aligned with [79] for emerging countries
Variables Level First difference
[80], for sub-Saharan Africa countries [81], for BRICS economies. As
Lnco2 0.882 (0.189) 4.992*** (0.000)
emerging economies like India, China, Argentina etc. are witnessing
Lnren 0.473 (0.318) 5.530*** (0.000)
Lngdp 0.069 (0.473) 3.593*** (0.000) a boom in internet penetration, the higher energy savings due to
Lngdpsq 0.050 (0.480) 3.420*** (0.000) the substitution effect of ICT is surpassing the rebound effect and
Lnfdi 0.144* (0.073) 4.560*** (0.000) compensation effect of the additional energy demand of ICT
Lnurban 7.165 (0.8600) 2.506*** (0.000) adoption. Moreover, rising internet penetration encourages
Lninternet 1.537*** (0.006) 3.450*** (0.000)
knowledge sharing and collaborative research with higher
461
A. Awan, K.R. Abbasi, S. Rej et al. Renewable Energy 189 (2022) 454e466

Table 5
Panel estimations through FMOL and DOLS.

Variables Dynamic OLS Fully Modified OLS

Coef S.E T-value Coef S.E T-value

Lnren 0.291*** 0.039117 7.439700 0.2564*** 0.009042 8.9032

Lngdp 2.129*** 0.679635 3.132926 1.8059*** 0.338631 5.3331
Lngdpsq 0.106*** 0.038101 2.784618 0.0857*** 0.018860 4.5446
Lnfdi 0.008* 0.011230 0.752823 0.0150* 0.006830 1.6620
Lnurban 1.244*** 0.453905 2.742570 0.8110*** 0.028803 3.8989
Lninternet 0.037*** 0.008855 3.927561 0.0170*** 0.008011 2.4992
R square 0.92
Adj- R square 0.91

Note: the sign (*) denotes the significance level, * at 10%, ** at5%, and *** at 1% level of significance.

Table 6 media rather than printed books or newspapers, etc. All these
Driscol Karaay Regression results. micro-level changes due to ICT are ultimately resulting in less
Variables (1) (2) automotive emission and reduced anthropogenic ecological foot-
prints at the global level.
lnCO2 lnCO2
Amongst other variables, the coefficient results reveal that GDP
lnren 0.258*** 0.263***
per capita increases on environmental degradation in the panel of
(0.0138) (0.0157)
lngdp 1.785*** 1.770*** 10 emerging countries. It shows that initially with rapid industri-
(0.195) (0.153) alization combined with accelerated economic growth environ-
lngdpsq 0.0847*** 0.0887*** mental deterioration is accelerated due to scale effect [82].
(0.0107) (0.00829) According to the EKC hypothesis, economic growth contributes
lnurban 0.818*** 0.783***
towards higher emissions as an economy becomes more industri-
(0.159) (0.121)
lnfdi 0.0123*** 0.00715* alized. Industrialization, in particular, necessitates a large-scale
(0.00415) (0.00404) consumption of energy resources, which results in significant
lninternet 0.0167*** environmental damage. Further, the GDPsq per capita result shows
(0.00450)
a negative influence of economic growth on emission, highlighting
Constant 10.28*** 9.721***
(0.326) (0.446)
that the adoption of an environmentally sustainable economic
growth model at a higher per capita income level helps in the
Observations 196 196
advancement of the technique effect. With consistent economic
R-sq 91.01 90.32
growth, the economy enters a post-industrialization period, during
Note: the Driscoll Kraay standard errors are stated in parentheses, ***p < 0.01,
which it reduces its emissions as a result of its economic structure
**p < 0.05, *p < 0.1.
and stringent environmental laws and regulations [83,84]. The
inverted U-shape EKC relationship results in this study are aligned
technological adoption, thereby facilitating the production of with the prior studies of [85]for Pakistan [86], for 14 Asian coun-
further energy-efficient devices. Interestingly, emerging economies tries [87], for Turkey [88], for South Asian [90], for China [89], for
such as China, Mexico, India, Turkey etc. are gradually becoming South Asian economies among others.
major hubs of production of communication devices and further The results of urbanization suggest a positive influence on CO2e,
increasing the pace of ICT adoption. With the widespread adoption which is aligned with the results of [91], for West Africa [92], for
of ICT, emerging economies are witnessing major overhauls in Asian economies among others. The extant literature highlights
lifestyle, such as working from home, online shopping, using digital that urbanization promotes non-renewable energy consumption

Table 7
Panel Quantile regression results.

Variables (1) (2) (3) (4) (5) (6) (7)

location scale qtile_5 qtile_25 qtile_50 qtile_75 qtile_95

lnren 0.2583*** 0.0028 0.2640*** 0.2609*** 0.2585*** 0.2559*** 0.2529***

(0.0193) (0.0111) (0.0314) (0.0228) (0.0194) (0.0208) (0.0279)
lngdp 1.7847*** 0.0011 1.7823*** 1.7836*** 1.7846*** 1.7856*** 1.7869***
(0.1825) (0.1048) (0.2965) (0.2153) (0.1834) (0.1969) (0.2634)
Lngdpsq 0.0847*** 0.0014 0.0877*** 0.0861*** 0.0848*** 0.0835*** 0.0820***
(0.0100) (0.0058) (0.0163) (0.0119) (0.0101) (0.0108) (0.0145)
lnfdi 0.0123* 0.0038 0.0202* 0.0159* 0.0125* 0.0089 0.0048
(0.0072) (0.0041) (0.0117) (0.0085) (0.0072) (0.0078) (0.0104)
lnurban 0.8183*** 0.0039 0.8102*** 0.8146*** 0.8180*** 0.8217*** 0.8259***
(0.1320) (0.0758) (0.2145) (0.1557) (0.1326) (0.1424) (0.1906)
lninternet 0.0167*** 0.0044 0.0076 0.0125** 0.0164*** 0.0206*** 0.0254***
(0.0049) (0.0028) (0.0081) (0.0058) (0.0049) (0.0053) (0.0071)
Constant 10.2796*** 0.0884 10.0954*** 10.1957*** 10.2730*** 10.3566*** 10.4536***
(0.6678) (0.3836) (1.0852) (0.7878) (0.6709) (0.7205) (0.9639)
Observations 196 196 196 196 196 196 196

Note: Standard errors are presented in parentheses, ***p < 0.01, **p < 0.05, *p < 0.1.

462
A. Awan, K.R. Abbasi, S. Rej et al. Renewable Energy 189 (2022) 454e466

Fig. 7. MMQREG and OLS coefficient at various quantile levels, yellow lines show MMQR and solid black line show OLS based coefficients.

[93], and industrialization, with an increase in automotive emis- levels, the study used the novel method of moments regression
sions, thereby contributing positively towards carbon emissions. technique formulated by Machado and Silva [28]. Furthermore,
Importantly, the results from all the regression models show a Fig. 2 shows the variation in coefficients of MMQREG across the
negative relationship between renewable energy and CO2e. The conditional distribution of CO2e emissions quantiles that shows
results are consistent with the results of [25,84] for BRICS countries supremacy over the traditional regression method. In addition, the
[32], for China [94e96], for a panel studies, while [97] for China figure shows that coefficient of renewable energy increases while
[98,99], for Pakistan among others. By substituting clean energy moving from 0.05th-0.90th quantile. This shows that renewable
supplies for fossil fuel energy, renewable energy helps to reduce energy is stronger in combating environmental damage when
GHG emissions. As a result, the development of renewable energy is carbon dioxide emission is at a lower level. The coefficient of GDP is
critical for emerging economies to reduce emissions. Moreover, ICT similar at all levels of quantile distribution; however, the squared
acts as a necessary enabler in this process by improving the effi- term is highly effective in reducing CO2e at lower quantile levels.
ciency of renewable energy devices through applications of artifi- This shows, that once the turning point of the EKC is achieved, a
cial intelligence and self-optimization facilities. very rapid reduction in environmental degradation is observed
which gradually reduces. The coefficient of FDI obtained through
5. Conclusion the method of moments regression shows that it raises CO2e
however, the magnitude of the coefficients tends to reduce as we
The goal of the present study is to analyze the long-run impacts move from 0.05th-0.50th quantile. Urbanization shows a consis-
of information communication technology, renewable energy, and tently positive effect on CO2e at the cross all quantile levels. This
foreign direct investment within the framework of the EKC hy- shows that regardless of the level of CO2e, emerging economies are
pothesis. In doing so, the study utilized balanced panel data for damaging the environment by urbanization. Lastly, our main
emerging countries during 1996e2015. The study examined data focused variable internet penetration, internet use the proxy for ICT
properties through the panel unit root and panel cointegration. Due shows a negative coefficient. Interestingly, the coefficient is 0.007
to the presence of cross-sectional dependency, a second-generation which increases to become 0.02 while we move from 0.05th to
unit root test for panel data is preferred. Regression analysis for 0.95th quantile levels of CO2e. Internet brings awareness about the
long-run relationships obtained through FMOLS, DOLS and Driscoll environment; it brings efficiency in all production process along
Kraay method supported the presence of the EKC in the sample of with reducing energy demand. Education and meetings are going
10 emerging countries. For a robust analysis of the impact of online saving a lot of energy in transportation. Similarly, due to
exogenous variables on CO2 emissions across various quantile online commerce, the traveling and printing of papers reduce and

463
A. Awan, K.R. Abbasi, S. Rej et al. Renewable Energy 189 (2022) 454e466

are helpful in environmental improvement. Table A2

Correlation Matrix

5.1. Policy implications Variables (1) (2) (3) (4) (5) (6) (7)

(1) lnco2 1.000

From a policy standpoint, the following steps might be taken to (2) lnren 0.869 1.000
(3) lngdp 0.729 0.687 1.000
achieve low-carbon society. The present study proposes that
(4) lngdpsq 0.713 0.679 0.999 1.000
renewable energy should be encouraged in emerging countries. (5) lnfdi 0.307 0.264 0.214 0.207 1.000
Along with this, ICT usage, particularly internet penetration should (6) lnurban 0.664 0.734 0.870 0.861 0.159 1.000
be supported. To initiate with, innovations that cut energy use are (7) lninternet 0.414 0.270 0.584 0.587 0.279 0.405 1.000
eagerly anticipated. As a result, regulations aimed at boosting en-
ergy efficiency in the ICT industry, as well as in household and
public buildings (such as the usage of smart appliances and smart List of abbreviations
grid ideas) might aid in reducing environmental deterioration.
Furthermore, increasing renewable energy supply and its per- EKC Environmental Kuznets Curve
centage of overall energy usage, as well as transitioning from short MMQREG Methods of Moments Quantile Regression
life cycle products and old equipment to contemporary equipment, ICT Information Communication Technologies
might lessen ICT's indirect influence on CO2 emissions. Policy- NARDL Nonlinear Autoregressive Distributive Lag
makers should encourage the emergence of green solutions by FDI Foreign Direct Investment
lowering the funding costs of environmentally friendly technology WDI World Development Indicators
and projects. These rewards should take the form of green subsidies FMOLS Fully Modified Ordinary Least Square
for the development and/or implementation of technologies. Any DOLS Dynamic Ordinary Least Square
tax on internet usage should be either reduced or completely be CO2e Carbon dioxide emissions
eliminated to enhance its usage. For better environmental protec- GDPsq Gross Domestic Product square
tion, internet usage must get government as well as private sector GHG Green House Gas
support. Finally, initiatives to raise public awareness about the
advantages and environmental costs of rising ICT penetration in
References
emerging countries should adhere to an integrated policy frame-
work that takes into account the stage of economic development in [1] K. Abbasi, F.F. Adedoyin, J. Abbas, K. Hussain, The impact of energy depletion
terms of urbanization and internet penetration level. and renewable energy on CO2 emissions in Thailand: fresh evidence from the
novel dynamic ARDL simulation, Renew. Energy 180 (2021) 1e12.
One of the study's drawbacks is that the results and policy
[2] K.R. Abbasi, Economic complexity, tourism, energy prices, and environmental
suggestions apply only to emerging nations and do not take into degradation in the top economic complexity countries: fresh panel evidence,
consideration the unique characteristics of each country. In truth, Environ. Sci. Pollut. Res. (2021).
[3] K.R. Abbasi, F. Fatai, A. Magdalena, The role of forest and agriculture towards
the growth patterns of ICT nations vary somewhat. As a result, it is
environmental fortification: designing a sustainable policy framework for top
critical to expanding this analysis to the national level to get a forested countries, Environ. Dev. Sustain. (2021), https://doi.org/10.1007/
better understanding of the effect of these policies. Future research s10668-021-01803-4.
might expand on this work by finding other transmission mecha- [4] N. Iqbal, K. Raza, R. Shinwari, W. Guangcai, M. Ahmad, K. Tang, Does exports
diversification and environmental innovation achieve carbon neutrality target
nisms, especially contextual ones, via which ICT penetration affects of OECD economies? J. Environ. Manag. 291 (2021) 112648.
environmental consequences other than CO2. [5] K. Raza, K. Hussain, M. Redulescu, I. Ozturk, Does natural resources depletion
and economic growth achieve the carbon neutrality target of the UK? A way
forward towards sustainable development, Resour. Policy 74 (2021),
CRediT authorship contribution statement 102e341.
[6] F. Ahmed, M. Naeem, M. Iqbal, ICT and renewable energy: a way forward to
the next generation telecom base stations, Telecommun. Syst. 64 (1) (2017)
Ashar Awan: Conceptualization, Methodology, Formal analysis, 43e56.
Software, Data curation. Kashif Raza Abbasi: Formal analysis, [7] A. Azam, M. Rafiq, M. Shafique, J. Yuan, An empirical analysis of the non-linear
Writing e original draft. Soumen Rej: Software, Validation, Writing effects of natural gas, nuclear energy, renewable energy and ICT-Trade in
leading CO2 emitter countries: policy towards CO2 mitigation and economic
e review & editing. Arunava Bandyopadhyay: Visualization, sustainability, J. Environ. Manag. 286 (2021) 112232, https://doi.org/10.1016/
Investigation, Resources, Data curation. Kangjuan Lv: Supervision. j.jenvman.2021.112232.
[8] M. Murshed, An empirical analysis of the non-linear impacts of ICT-trade
openness on renewable energy transition, energy efficiency, clean cooking
Declaration of competing interest fuel access and environmental sustainability in South Asia, Environ. Sci. Pollut.
Control Ser. 27 (29) (2020) 36254e36281. https://10.1007/s11356-020-
09497-3.
The authors declare that they have no known competing [9] M. Murshed, M.H. Chadni, J. Ferdaus, Does ICT trade facilitate renewable en-
financial interests or personal relationships that could have ergy transition and environmental sustainability? Evidence from Bangladesh,
appeared to influence the work reported in this paper. India, Pakistan, Sri Lanka, Nepal and Maldives, Energy, Ecol. Environ. 5 (6)
(2020) 470e495. https://10.1007/s40974-020-00190-2.
[10] N. Koshta, H.A. Bashir, T.A. Samad, Foreign trade, financial development,
Appendix agriculture, energy consumption and CO2 emission: testing EKC among
emerging economies, Indian Growth Dev. Rev. 14 (1) (2021) 50e80. https://
10.1108/IGDR-10-2019-0117.
[11] M.W. Zafar, F.M. Mirza, S.A.H. Zaidi, F. Hou, The nexus of renewable and
nonrenewable energy consumption, trade openness, and CO 2 emissions in
Table A1 the framework of EKC: evidence from emerging economies, Environ. Sci.
Cointegration Test (Kao Residual) Pollut. Control Ser. 26 (15) (2019) 15162e15173.
[12] S.P. Dimelis, S.K. Papaioannou, ICT growth effects at the industry level: a
comparison between the US and the EU, Inf. Econ. Pol. 23 (1) (2011) 37e50.
ADF t-Statistic Prob. [13] M. O'Mahony, M. Vecchi, Quantifying the impact of ICT capital on output
4.264170 0.0000 growth: a heterogeneous dynamic panel approach, Economica 72 (288)
Residual variance 0.001287 (2005) 615e633.
[14] Danish, B. Zhang, B. Wang, Z. Wang, Role of renewable energy and non-
HAC variance 0.001072
renewable energy consumption on EKC: evidence from Pakistan, J. Clean.

464
A. Awan, K.R. Abbasi, S. Rej et al. Renewable Energy 189 (2022) 454e466

Prod. 156 (2017) 855e864. https://10.1016/j.jclepro.2017.03.203. [42] H.E. Moghadam, V.D. Vahid, The impact of financial development and trade on
[15] L.M. N’dri, M. Islam, M. Kakinaka, ICT and environmental sustainability: any environmental quality in Iran, Empir. Econ. 54 (2018) 1777e1799, https://
differences in developing countries? J. Clean. Prod. 297 (2021) 126642. doi.org/10.1007/s00181-017-1266-x.
https://10.1016/j.jclepro.2021.126642. [43] I. Ozturk, U. Al-mulali, Investigating the validity of the environmental Kuznets
[16] Simplice A. Asongu, ICT, openness and CO2 emissions in Africa, Environ. Sci. curve hypothesis in Cambodia, Ecol. Indicat. 57 (2015) 324e330, https://
Pollut. Control Ser. 1 (1) (2018) 9351e9359. https://10.23959/sfahrj-1000006. doi.org/10.1016/j.ecolind.2015.05.018.
[17] U. Al-Mulali, L. Sheau-Ting, I. Ozturk, The global move toward Internet [44] D. Pal, S. Kumar, The environmental Kuznets curve for carbon dioxide in India
shopping and its influence on pollution: an empirical analysis, Environ. Sci. and China : growth aChina:ution at crossroad, J. Pol. Model. 39 (2017)
Pollut. Control Ser. 22 (13) (2015) 9717e9727. https://10.1007/s11356-015- 371e385, https://doi.org/10.1016/j.jpolmod.2017.03.005.
4142-2. [45] S. Pandey, M. Mishra, CO2 emissions and economic growth of SAARC Coun-
[18] V.C. Coroama, L.M. Hilty, M. Birtel, Effects of Internet-based multiple-site tries: evidence from a panel var analysis, World J. Appl. Econ. 1 (2015) 23e33.
conferences on greenhouse gas emissions, Telematics Inf. 29 (4) (2012) [46] M.M. Rahman, R. Nepal, K. Alam, Impacts of human capital, exports, economic
362e374. https://10.1016/j.tele.2011.11.006. growth and energy consumption on CO2 emissions of a cross-sectionally
[19] S.A. Asongu, S. Le Roux, Enhancing ICT for inclusive human development in dependent panel: evidence from the newly industrialized countries (NICs),
Sub-Saharan Africa, Technol. Forecast. Soc. Change 118 (2017) 44e54, https:// Environ. Sci. Pol. 121 (2021) 24e36, https://doi.org/10.1016/
doi.org/10.1016/j.techfore.2017.01.026. j.envsci.2021.03.017.
[20] Y. Park, F. Meng, M.A. Baloch, The effect of ICT, financial development, growth, [47] M. Ur, M. Rashid, Energy consumption to environmental degradation, the
and trade openness on CO2 emissions: an empirical analysis, Environ. Sci. growth appetite in SAARC nations, Renew. Energy 111 (2017) 284e294,
Pollut. Control Ser. 25 (30) (2018) 30708e30719. https://10.1007/s11356- https://doi.org/10.1016/j.renene.2017.03.100.
018-3108-6. [48] K. Dong, G. Hochman, Y. Zhang, R. Sun, H. Li, H. Liao, CO2 emissions, economic
[21] P. Sadorsky, The effect of urbanization on CO2 emissions in emerging econ- and population growth, and renewable energy: empirical evidence across
omies, Energy Econ. 41 (2014) 147e153, https://doi.org/10.1016/ regions, Energy Econ. 75 (2018) 180e192.
j.eneco.2013.11.007. [49] E. Koçak, A. Şarkgüneşi, The renewable energy and economic growth nexus in
[22] D. Wang, T. Zhou, M. Wang, Information and communication technology (ICT), Black Sea and Balkan countries, Energy Pol. 100 (2017) 51e57, https://doi.org/
digital divide and urbanization: evidence from Chinese cities, Technol. Soc. 64 10.1016/j.enpol.2016.10.007.
(2021) 101516, https://doi.org/10.1016/j.techsoc.2020.101516. [50] S. Kais, H. Sami, An econometric study of the impact of economic growth and
[23] A. Kasman, Y.S. Duman, CO2 emissions, economic growth, energy consump- energy use on carbon emissions: panel data evidence from fifty-eight coun-
tion, trade and urbanization in new EU member and candidate countries: a tries, Renew. Sustain. Energy Rev. 59 (2016) 1101e1110.
panel data analysis, Econ. Modell. 44 (2015) 97e103, https://doi.org/10.1016/ [51] A.N. Menegaki, Growth and renewable energy in Europe: a random effect
j.econmod.2014.10.022. model with evidence for neutrality hypothesis, Energy Econ. 33 (2011),
[24] R.P. Pradhan, M.B. Arvin, M. Nair, Urbanization, transportation infrastructure, https://doi.org/10.1016/j.eneco.2010.10.004.
ICT, and economic growth: a temporal causal analysis, Cities 115 (2021) [52] N. Doytch, S. Narayan, Does FDI influence renewable energy consumption? An
103213, https://doi.org/10.1016/j.cities.2021.103213. analysis of sectoral FDI impact on renewable and non-renewable industrial
[25] Danish, N. Khan, M.A. Baloch, S. Saud, T. Fatima, The effect of ICT on CO2 energy consumption, Energy Econ. 54 (2016) 291e301.
emissions in emerging economies: does the level of income matters? Environ. [53] P. Sadorsky, The effect of urbanization on CO2 emissions in emerging econ-
Sci. Pollut. Control Ser. 25 (23) (2018) 22850e22860. https://10.1007/s11356- omies, Energy Econ. 41 (2014) 147e153, https://doi.org/10.1016/
018-2379-2. j.eneco.2013.11.007.
[26] H.G. Fereidouni, Foreign direct investments in real estate sector and CO2 [54] P. Poumanyvong, S. Kaneko, Does urbanization lead to less energy use and
emission: evidence from emerging economies, Manag. Environ. Qual. Int. J. 24 lower CO2 emissions? A cross-country analysis, Ecol. Econ. 70 (2010)
(4) (2013) 463e476. https://10.1108/MEQ-04-2012-0032. 434e444, https://doi.org/10.1016/j.ecolecon.2010.09.029.
[27] B. Liddle, What are the carbon emissions elasticities for income and popula- [55] R.L. Moomaw, A.M. Shatter, Urbanization and economic development: a bias
tion? Bridging STIRPAT and EKC via robust heterogeneous panel estimates, toward large cities? J. Urban Econ. 40 (1996) 13e37.
Global Environ. Change 31 (2015) 62e73, https://doi.org/10.1016/ [56] O.U. Oro, I.P. Alagidede, Does petroleum resources or market size drive foreign
j.gloenvcha.2014.10.016. direct investment in Africa? New evidence from time-series analysis, Resour.
[28] J.A.F. Machado, J.M.C. Santos Silva, Quantiles via moments, J. Econom. 213 (1) Policy 71 (2021) 101992.
(2019) 145e173. https://10.1016/j.jeconom.2019.04.009. [57] L. Zarsky, Havens, halos and spaghetti: untangling the relationship between
[29] G.M. Grossman, A.B. Krueger, Environmental impacts of a north American free FDI and the environment, Foreign direct Invest. Environ. (1999) 47e73.
trade agreement, Natl. Bur. Econ. Res. Work. Pap. Ser. No. 3914 (1991) 1e57, [58] M.A. Cole, Trade, the pollution haven hypothesis and the environmental
https://doi.org/10.3386/w3914. Kuznets curve: examining the linkages, Ecol. Econ. 48 (2004) 71e81.
[30] S. Kuznets, This content downloaded from 193.255.88.109 on Sun, Am. Econ. [59] K. Shehzad, L. Xiaoxing, M. Sarfraz, Envisaging the asymmetrical association
Rev. 45 (1955) 32. among FDI, ICT, and climate change: a case from developing country, Carbon
[31] F. Bilgili, S. Kuşkaya, M. Khan, A. Awan, O. Türker, The roles of economic Manag. 12 (2021) 123e137, https://doi.org/10.1080/17583004.2021.1890449.
growth and health expenditure on CO2 emissions in selected Asian countries: [60] P. Jokinen, P. Malaska, J. Kaivo-Oja, The environment in aninformation soci-
a quantile regression model approach, Environ. Sci. Pollut. Res. (2021), https:// ety’: a transition stage towards more sustainable development? Futures 30
doi.org/10.1007/s11356-021-13639-6. (1998) 485e498.
[32] K. Dong, R. Sun, H. Jiang, X. Zeng, CO2 emissions, economic growth, and the [61] N. Roome, J. Park, Introduction: global sustainability and information econ-
environmental Kuznets curve in China: what roles can nuclear energy and omy: old challenges, new perspectives, Greener Manag. Int. (2000) 24e32.
renewable energy play? J. Clean. Prod. 196 (2018) 51e63. [62] D.Z. Sui, D.W. Rejeski, Environmental impacts of the emerging digital econ-
[33] O. Fernandez-Amador, J.F. Francois, D.A. Oberdabernig, P. Tomberger, Carbon omy: the e-for-environment e-commerce? Environ. Manage. 29 (2002)
dioxide emissions and economic growth: an assessment based on production 155e163.
and consumption emission inventories, Ecol. Econ. 135 (2017) 269e279. [63] M.W. Toffel, A. Horvath, Environmental implications of wireless technologies:
[34] A. Hashemizadeh, Q. Bui, N. Kongbuamai, Unpacking the role of public debt in news delivery and business meetings, Environ. Sci. Technol. 38 (2004)
renewable energy consumption: new insights from the emerging countries, 2961e2970.
Energy 224 (2021) 120187, https://doi.org/10.1016/j.energy.2021.120187. [64] L. Erdmann, L.M. Hilty, Scenario analysis: exploring the macroeconomic im-
[35] J.I. Mikayilov, M. Galeotti, F.J. Hasanov, The impact of economic growth on pacts of information and communication technologies on greenhouse gas
CO2 emissions in Azerbaijan, J. Clean. Prod. 197 (2018) 1558e1572, https:// emissions, J. Ind. Ecol. 14 (2010) 826e843.
doi.org/10.1016/j.jclepro.2018.06.269. [65] J.D. Moyer, B.B. Hughes, ICTs: do they contribute to increased carbon emis-
[36] H. Fujii, S. Managi, Which industry is greener? An empirical study of nine sions? Technol. Forecast. Soc. Change 79 (2012) 919e931.
industries in OECD countries, Energy Pol. 57 (2013) 381e388. [66] M. Salahuddin, K. Alam, I. Ozturk, The effects of Internet usage and economic
[37] K. Zaman, M. Abd-el Moemen, Energy consumption, carbon dioxide emissions growth on CO2 emissions in OECD countries: a panel investigation, Renew.
and economic development: evaluating alternative and plausible environ- Sustain. Energy Rev. 62 (2016) 1226e1235, https://doi.org/10.1016/
mental hypothesis for sustainable growth, Renew. Sustain. Energy Rev. 74 j.rser.2016.04.018.
(2017) 1119e1130. [67] I. Røpke, T.H. Christensen, Energy impacts of ICTeInsights from an everyday
[38] A.N. Ajmi, S. Hammoudeh, D.K. Nguyen, J.R. Sato, On the relationships be- life perspective, Telematics Inf. 29 (2012) 348e361.
tween CO2 emissions, energy consumption and income: the importance of [68] H. Ishida, The effect of ICT development on economic growth and energy
time variation, Energy Econ. 49 (2015) 629e638, https://doi.org/10.1016/ consumption in Japan, Telematics Inf. 32 (2015) 79e88.
j.eneco.2015.02.007. [69] J.L. Funk, IT and sustainability: new strategies for reducing carbon emissions
[39] B. Dong, F. Wang, Y. Guo, The global EKCs, Int. Rev. Econ. Finance 43 (2016) and resource usage in transportation, Telecommun. Pol. 39 (2015) 861e874.
210e221, https://doi.org/10.1016/j.iref.2016.02.010. [70] U. Al-Mulali, L. Sheau-Ting, I. Ozturk, The global move toward Internet
[40] M. Javid, M. Khan, Energy efficiency and underlying carbon emission trends, shopping and its influence on pollution: an empirical analysis, Environ. Sci.
Environ. Sci. Pollut. Res. 27 (2020) 3224e3236, https://doi.org/10.1007/ Pollut. Res. 22 (2015) 9717e9727.
s11356-019-07019-4. [71] H. Hamdi, R. Sbia, M. Shahbaz, The nexus between electricity consumption
[41] J.I. Mikayilov, F.J. Hasanov, Emissions-economic growth relationship revisited: and economic growth in Bahrain, Econ. Modell. 38 (2014) 227e237.
new insights from the time-varying cointegration approach, STATISTIKA 100 [72] M. Salahuddin, K. Alam, Internet usage, electricity consumption and economic
(2020) 2. growth in Australia: a time series evidence, Telematics Inf. 32 (2015)

465
A. Awan, K.R. Abbasi, S. Rej et al. Renewable Energy 189 (2022) 454e466

862e878. structural breaks, J. Clean. Prod. 187 (2018) 770e779, https://doi.org/10.1016/

[73] R. Koenker, Quantile regression for longitudinal data, J. Multivariate Anal. 91 j.jclepro.2018.03.236.
(1) (2004). [88] M. Murshed, LPG consumption and environmental Kuznets curve hypothesis
[74] I.A. Canay, A simple approach to quantile regression for panel data, Econom. J. in South Asia: a time-series ARDL analysis with multiple structural breaks,
14 (3) (2011) 368e386. Environ. Sci. Pollut. Res. (2020), https://doi.org/10.1007/s11356-020-10701-7.
[75] Y. Shin, B. Yu, M. Greenwood-Nimmo, Modelling asymmetric cointegration [89] M. Murshed, N.T.T. Dao, Revisiting the CO2 Emission-Induced EKC Hypothesis
and dynamic multipliers in a nonlinear ARDL framework, in: R. Sickles, in South Asia: the Role of Export Quality Improvement, GeoJournal, 2020,
W. Horrace (Eds.), Festschrift in Honor of Peter Schmidt, Springer, New York, https://doi.org/10.1007/s10708-020-10270-9.
NY, 2014, https://doi.org/10.1007/978-1-4899-8008-3_9. [90] T. Li, Y. Wang, D. Zhao, Environmental Kuznets Curve in China: new evidence
[76] M.H. Pesaran, A simple panel unit root test in the presence of cross-section from dynamic panel analysis, Energy Pol. 91 (2016) 138e147, https://doi.org/
dependence, J. Appl. Econom. 22 (2007) 265e312, https://doi.org/10.1002/ 10.1016/j.enpol.2016.01.002.
jae.951. [91] M. Musah, Y. Kong, I.A. Mensah, et al., The connection between urbanization
[77] J. Westerlund, Testing for error correction in panel data, Oxf. Bull. Econ. Stat. and carbon emissions: a panel evidence from West Africa, Environ. Dev.
69 (2007) 709e748, https://doi.org/10.1111/j.1468-0084.2007.00477.x. Sustain. 23 (2021) 11525e11552, https://doi.org/10.1007/s10668-020-01124-
[78] C. Kao, Spurious regression and residual-based tests for cointegration in panel y.
data, J. Econom. 90 (1999) 1e44, https://doi.org/10.1016/S0304-4076(98) [92] R. Salim, S. Rafiq, S. Shafiei, Y. Yao, Does urbanization increase pollutant
00023-2. emission and energy intensity? evidence from some Asian developing econ-
[79] B. Ozcan, N. Apergis, The impact of internet use on air pollution: evidence omies, Appl. Econ. 51 (2019) 4008e4024, https://doi.org/10.1080/
from emerging countries, Environ. Sci. Pollut. Res. 25 (2018) 4174e4189, 00036846.2019.1588947.
https://doi.org/10.1007/s11356-017-0825-1. [93] R.A. Salim, S. Shafiei, Urbanization and renewable and non-renewable energy
[80] S.A. Asongu, ICT, openness and CO 2 emissions in Africa, Environ. Sci. Pollut. consumption in OECD countries: an empirical analysis, Econ. Modell. 38
Control Ser. 25 (10) (2018) 9351e9359. (2014) 581e591, https://doi.org/10.1016/j.econmod.2014.02.008.
[81] A. Haseeb, E. Xia, S. Saud, et al., Does information and communication tech- [94] A. Allard, J. Takman, G.S. Uddin, A. Ahmed, The N-shaped environmental
nologies improve environmental quality in the era of globalization? An Kuznets curve: an empirical evaluation using a panel quantile regression
empirical analysis, Environ. Sci. Pollut. Res. 26 (2019) 8594e8608, https:// approach, Environ. Sci. Pollut. Res. 25 (2018) 5848e5861, https://doi.org/
doi.org/10.1007/s11356-019-04296-x. 10.1007/s11356-017-0907-0.
[82] A. Musolesi, M. Mazzanti, R. Zoboli, A panel data heterogeneous Bayesian [95] G. Aydin, I. Karakurt, K. Aydiner, Analysis and mitigation opportunities of
estimation of environmental Kuznets curves for CO2 emissions, Appl. Econ. 42 methane emissions from the energy sector. Energy sources, Part A recover,
(2010) 2275e2287, https://doi.org/10.1080/00036840701858034. Util. Environ. Eff. 34 (2012) 967e982, https://doi.org/10.1080/
[83] A. Bandyopadhyay, S. Rej, Can nuclear energy fuel an environmentally sus- 15567031003716725.
tainable economic growth? Revisiting the EKC hypothesis for India, Environ. [96] G. Aydin, The application of trend analysis for coal demand modeling. Energy
Sci. Pollut. Res. (2021), https://doi.org/10.1007/s11356-021-15220-7. sources, Part B econ, Planning, Policy 10 (2015) 183e191, https://doi.org/
[84] C. Cheng, X. Ren, Z. Wang, C. Yan, Heterogeneous impacts of renewable energy 10.1080/15567249.2013.813611.
and environmental patents on CO2 emission - evidence from the BRIICS, Sci. [97] K.R. Abbasi, M. Shahbaz, J. Zhang, M. Irfan, K. Lv, Analyze the Environmental
Total Environ. 668 (2019) 1328e1338, https://doi.org/10.1016/ Sustainability Factors of China: the Role of Fossil Fuel Energy and Renewable
j.scitotenv.2019.02.063. Energy, Renewable Energy, 2022.
[85] M. Nasir, F.U. Rehman, Environmental Kuznets curve for carbon emissions in [98] K.R. Abbasi, K. Hussain, A.M. Haddad, A. Salman, I. Ozturk, The role of financial
Pakistan: an empirical investigation, Energy Pol. 39 (2011) 1857e1864. development and technological innovation towards sustainable development
[86] N. Apergis, I. Ozturk, Testing environmental Kuznets curve hypothesis in in Pakistan: fresh insights from consumption and territory-based emissions,
Asian countries, Ecol. Indicat. 52 (2015) 16e22, https://doi.org/10.1016/ Technol. Forecast. Soc. Change 176 (2022) 121444.
j.ecolind.2014.11.026. [99] K.R. Abbasi, M. Shahbaz, Z. Jiao, M. Tufail, How energy consumption, industrial
[87] U.K. Pata, Renewable energy consumption, urbanization, financial develop- growth, urbanization, and CO2 emissions affect economic growth in Pakistan?
ment, income and CO2 emissions in Turkey: testing EKC hypothesis with A novel dynamic ARDL simulations approach, Energy 221 (2021) 119793.

466