1 s2.0 S1342937X23000369 Main

Gondwana Research 119 (2023) 27–44
Contents lists available at ScienceDirect
Gondwana Research
journal homepage: www.elsevier.com/locate/gr
Effect of wind and solar energy production, and economic development

on the environmental quality: Is this the solution to climate change?
Bilal Hussain, Syed Asif Ali Naqvi ⇑, Sofia Anwar, Muhammad Usman
Department of Economics, Government College University, Faisalabad 38000, Pakistan
a r t i c l e i n f o a b s t r a c t
Article history: The motivation of the present study is to evaluate the effect of wind and solar energy production and eco-
Received 20 April 2022 nomic development on environmental quality in 43 countries covering the period of 1990–2017. The
Revised 18 December 2022 study uses annual data for real GDP, real trade, energy intensity, urbanization, foreign direct investment,
Accepted 27 January 2023
ecological footprints, and wind and solar energy production. For econometric analysis, common corre-
Available online 2 February 2023
Handling Editor: Sanghoon Kwon
lated mean group (CCE-MG), cross-section-distributive lag (CS-DL), and cross-section autoregressive-
distributive lag (CS-ARDL) approaches are employed. The outcomes show that real GDP has a significantly
positive influence on ecological footprint. At the same time, wind and solar energy negatively and signif-
Keywords:
icantly influence ecological footprint. It is argued that a 1 % increase in wind and solar energy reduces
Wind and solar energy ecological footprint by 3.1 % and 2.9 %, respectively. For policy recommendations, the study recom-
Economic development mended that wind and solar energy resources may become more noticeable for electricity purposes
Environment despite rising investments in these energy resources. Indeed, the share of wind and solar energy produc-
Ecological footprint tion is limited and dominated by non-renewable energy resources, so low-carbon energy resources might
Carbon emission be encouraged.
Renewable energy Ó 2023 International Association for Gondwana Research. Published by Elsevier B.V. All rights reserved.
1. Introduction the nations (Gullapalli, 2020). The structural change and paradigm
policies have reduced GHGs by shifting the consumption of non-
For more than two decades, the increasing threat of global renewable energy toward cleaner energy technologies. Conse-
warming and climate change has been a significant concern for quently, such advancement in technology, execution of environ-
mental protocols and environmental sustainability has declined
the levels of harmful pollutants against economic development in
Abbreviations: CO2, Carbon dioxide; GDP, Gross domestic product; EF, Ecological the developed world (Golpîra et al., 2018; Hao et al., 2018). Despite
footprint; FDI, Foreign direct investment; GHGs, Greenhouse gases; R&D, Research that, the mix of harmful pollutants has changed from GHGs to solid
and development; lnEF, Log of ecological footprint; lnURB, Log of urbanization; lnRT,
waste and effluents; as a result, the quantity of solid waste is still
Log of real trade; lnWS, Log of wind and solar energy; lnRGDP, Log of real GDP; lnEI,
Log of energy intensity; lnFDI, Log of foreign direct investment; EKC, Environment high per person.
kuznet curve; CCEMG, Common correlated effects mean group; D-H, Dumitrescu Environmental sustainability in developing and developed
and Hurlin causality test; CSD, Cross sectional dependence; VECM, Vector error nations depends entirely on GDP growth; accordingly, environ-
correction model; ARDL, Autoregressive distributive lag; AMG, Augmented mean
mental quality improves after a threshold level of GDP. Economic
group; DOLS, Dynamic ordinary least square; CUP-FMOLS, Continuously updated
fully modified OLS; BRICS, Brazil, Russia, India, China, South Africa; APEC, Asia-
growth and development are mainly connected with social growth,
Pacific Economic Cooperation countries; GCC, Gulf cooperation council; USA, United FDI, urbanization, foreign trade, and energy supply (Van & Bao,
states of America; UNIDO, United Nations Industrial Development Organization; 2018). Equally important, the critical role of the energy sector in
OECD, Organization for Economic Co-operation and Development; OLS, Ordinary economic development and a sustainable environment may not
least square; CIPS, Cross sectional augmented test; CADF, Cross- sectional
be ignored, as it is an essential input in the production process
augmented dickey fuller; FMOLS, Fully modified ordinary least square; DOLS,
Dynamic ordinary Least Square; ARDL, Autoregressive distributive lag model; IEA, (Mercure et al., 2018). Several studies examined the nexus among
International energy agency; WDI, World development indicators; DCCE CS-DL, energy, growth and carbon emissions (Azam et al., 2021; Wang
Dynamic common correlated effects cross section-distributive lag; DCCE CS-ARDL, et al., 2016; Zhang and Cheng, 2009; Mele et al., 2021;
Dynamic common correlated effects cross section autoregressive-distributive lag.
Magazzino et al. 2021; Magazzino et al. 2022). Most of these stud-
⇑ Corresponding author.
ies proved the negative association between energy consumption
E-mail addresses: bilalhussain1410@gmail.com (B. Hussain), syedasif_1@yahoo.
com (S. Asif Ali Naqvi), sofia_eco@gcuf.edu.pk (S. Anwar), usmanchoudhry786@g- and carbon emission for any given production unit, and this rela-
mail.com (M. Usman). tionship varied imperatively across the regions..
https://doi.org/10.1016/j.gr.2023.01.012
1342-937X/Ó 2023 International Association for Gondwana Research. Published by Elsevier B.V. All rights reserved.
B. Hussain, S. Asif Ali Naqvi, S. Anwar et al. Gondwana Research 119 (2023) 27–44
Usually, developing and developed nations utilize more energy lished with a similar context apart from Bulut and Menegaki
from conventional resources instead of cleaner energy technolo- (2020) for the top 10 nations with the highest installed capacity,
gies for economic activities. Using such non-renewable energy Magazzino (2012) for Italy, and Magazzino et al. (2020) for India,
resources further leads to ecological destruction and environmen- China and USA.
tal deficit in developed nations such as the USA, Brazil, Germany In brief, to our best knowledge, no single study evaluated the
and China etc. (Ahmed et al., 2020; Doytch, 2020; Destek and effect of wind and solar energy production and GDP on EF for these
Sinha, 2020). Due to such adverse environmental effects, con- nations of the world. Wind and solar energy reduce fossil combus-
cerned policymakers and stakeholders have begun to consider tion and provide air-quality and greenhouse gas emission benefits.
and promote alternative energy resources to reduce CO2 emissions These benefits vary dramatically by region and over time (Millstein
(Kuriqi et al., 2019). However, Ali et al. (2019) offered evidence of et al., 2017). Previous studies have only investigated renewable
the undesirable influence of hydropower energy on the ecosystem. energy benefits for the individual nation or multi-regional level
Further, nations need to raise energy production from cleaner by focusing on single regions. As a result, this study strives to
energy technologies. Rather than increasing the ratio of renewable add to the available literature with three innovative perspectives.
energy by installed capacity, it is also necessary for countries to Firstly, this study is the first to focus on a very inclusive dataset
emphasize the infrastructure for renewable energy (Chen et al., of 43 nations of the world, exploring the influence of renewable
2019). energy production and GDP on EF.
According to some authors (e.g. Goh and Ang, 2018), renewable Second, parallel to past studies, this study emphasized wind
energy is the best substitute for non-renewable energy and thus and solar energy production as a promising renewable energy
leads to less CO2 emissions. It is optimal for the structure of cleaner source and checked their role on environmental quality. This arti-
energy production and consumption. So, cleaner energy technolo- cle examines the possibility of wind and solar energy production to
gies have become an adequate substitute for conventional energy ensure environmental quality in favour of economic development.
and open the way toward a sustainable environment and growth Third, to achieve desired objectives, the study utilizes three inno-
(Bhattacharya et al., 2017). Bertram et al. (2015) suggested that vative approaches, CCE-MG, CS-DL and CS-ARDL. This choice is
low-carbon energy technologies paved the way for decarburization inspired by the remarks by Phillips and Sul (2003) and Paramati
in the future. The demand for cleaner energy technologies, such as et al. (2017), who remarked that conventional panel regression
solar, wind, biofuels, waste, hydropower, and biomass, has methods could be inconsistent and biased in the presence of CSD,
increased globally (Dong et al., 2017). Moreover, to fast-track and the CCE-MG regression method is robust to slope homogeneity
accessibility of renewable energy, UNO is emplaning the develop- and CSD developed by Pesaran (2006). In addition, Chudik and
ment and promotion of renewable energy resources as a business. Pesaran (2015) developed Dynamic common correlated effects
Business models for producing renewable energy will give inde- cross-section-distributive lag (DCCE CS-DL) and Dynamic common
pendent powers to producers, raise energy entrepreneurs, and con- correlated effects cross-section autoregressive-distributive lag
vert the agriculture industry into rural enterprises for energy approaches (DCCE CS-ARDL) to resolve the issue of CSD in panel
(UNIDO, 2019). data.
Hereafter, investigating the indicators of GHGs emissions as a In consequence, the contribution of current research to litera-
proxy for climate change may confine the exploration of industrial ture is empirical as well as methodological. Theoretically, its target
activities and urbanization. However, it ignores the negative influ- is to use an updated data testing approach to execute a compre-
ence of anthropogenic actions on the ecosystem in biodiversity and hensive causality valuation. It conveys new visions on the real
EF (Charfeddine, 2017; Danish and Wang, 2019). Compared to this influence of shifting energy regarding policies from non-
framework, the present study firmly highlights the EF as a crucial renewable energy to cleaner technologies worldwide. Moreover,
gauge of the ecosystem associated with resource utilization. Gen- the findings of this study are assumed to be fruitful for concerned
erally, EF represents the reformative natural capability needed policymakers and stakeholders in the decision-making process.
for the production process and consumption of resources by Finally, the present study inspects the effect of wind and solar
humans. Moreover, it promises an extensive choice of data regard- energy production and GDP for 43 countries with an installed wind
ing the environment into one indicator, and thus research on EF and solar energy capacity. Covering 1990–2017, this study
may provide new inferences to eliminate the ecological externali- assumes a panel data approach to give suitable policy suggestions
ties for the nations (Wang et al., 2020). for the world. CCE-MG, CS-DL and CS-ARDL approaches are applied
Against this background, some crucial gaps in the available lit- for the long-run findings. Also, the D-H test is utilized to observe
erature motivate this study. First, several studies investigated the the causality among the study variables.
nexus between growth, EF and renewable energy production. The paper is divided into six sections. Following the introduc-
Despite that, many studies performed single-country analysis, tion (section 1), the literature review occupies Section 2. Sec-
e.g., Ahmad and Du (2017) for Iran, Magazzino (2017) for Italy, tion 3 examines the methodology, while Sections 4 and 5 present
Danish et al. (2018) for Pakistan, Hdom and Fuinhas (2020) for Bra- the results, and discussion of the study. The last section provides
zil, Chen et al. (2019), Hu et al. (2014) and Rahman et al. (2020) for conclusions, policy recommendations and future research avenues.
China, Jaramillo and Muller (2016) for the USA and Blindheim
(2015) for Norway and Sweden. Consequently, these studies pro-
vided contradictory findings, but no single study was found on a 2. Brief literature review
large sample of countries in a single analysis. Regardless, a few
multi-country studies have been accomplished (see refs. Khan Since the last decade, a plethora of literature has been available
et al. (2019) for seven ASEAN nations, Ohler and Fetters (2014) on investigating environmental quality aspects. Researchers used
for 20 OECD countries, and Przychodzen and Przychodzen (2020) different economic and environmental quality variables and many
for 27 transition economies), but may not provide a clear theme used CO2 emissions as a proxy for environmental degradation. The
of findings on this topic. Second, the previous literature on the quantity of CO2 emissions may be determined by numerous social,
nexus among GDP, EF and renewable energy production at disag- economic, geographical, and political indicators. There is a short-
gregate levels (especially at the wind and solar energy) is minimal. age of literature on other indicators of the environment, such as
Almost it is cleared that no single article on the nexus of wind and EF. However, different authors used different variables of environ-
solar energy production, economic growth, and EF has been pub- ment quality to explore the dynamic associations. In addition, pre-
28
vious literature mostly used three kinds of datasets, cross-country, to 2012. Results established the feedback theory among GDP and
panel and time series, for examining the empirical relationships. trade openness, renewable energy to GDP and trade openness.
The results of studies were dissimilar in terms of direction and Dong et al. (2018) used a modern econometric approach known
magnitude of association between EF and numerous economic as CCE-MG to analyze the association between the population,
and energy variables. renewable energy intensity, GDP and CO2. The estimated findings
Our study objective is to evaluate the influence of wind and showed that economic growth and population increase CO2, and
solar energy production and GDP on EF. So, in this section, we tried renewable energy significantly reduced CO2.
to discuss the past literature about the effect of energy and GDP on Further, Jebli et al. (2020) found the adverse effect of renewable
the environment across numerous contexts. In this regard, the lit- energy consumption on industrial value-added worldwide. Using
erature is distributed in the following sub-headings: studies on CUP-FMOLS, Zafar et al. (2019) analyzed APEC economies and
energy, GDP and EF; studies on renewable energy, GDP and EF; dis- found energy use, trade, and R&D expenditures as crucial variables
aggregate studies on renewable energy, GDP and EF. for GDP. Another study by Sharif et al. (2020) identified the link
between renewable energy use and CO2 emission using the
2.1. Studies on energy, GDP and EF quintile-on-quintile regression method in the top 10 polluted
countries. The findings revealed the negative association between
Here is a bulk of literature that examined the link amongst CO2 emission and renewable energy use these countries. For China,
energy, environment quality and GDP (Pao et al., 2011; Al-mulali Chen et al. (2019) used ARDL and VECM methods and analyzed
and Sab, 2012; Cowan et al., 2014; Fei et al., 2011; Charfeddine, that renewable energy production lessens CO2, whereas non-
2017; Jammazi and Aloui, 2015; Wang et al., 2018; Raza et al., renewable energy increases CO2 emissions.
2019; Dabachi et al., 2020; Baz et al., 2020). For instance, Pao In contrast, Maji et al. (2019) used the ARDL model for 15 West
et al. (2011) explained that the EKC theory is not maintained in African economies. They estimated that renewable energy reduces
Russia and found that output negatively impacts CO2 emissions. productivity and GDP because of inefficient and unclean energy
Fei et al. (2011) observed the association between GDP, energy resources. Moreover, Shahbaz et al. (2020) and Rahman and
and the environment and argued that energy use positively affects Velayutham (2020) used the same methodology for empirical esti-
GDP in China. Similarly, Al-mulali and Sab (2012) studied 30 Sub- mation in the context of 38 non-renewable energy-using econo-
Saharan African economies and confirmed the positive influence of mies and 5 South Asian economies, respectively. Also, they
energy use on CO2, financial development and GDP. provided the same findings, e.g., gross fixed capital formation,
In addition, using the panel Granger causality test, Cowan et al. renewable and non-renewable energy use positively affect GDP
(2014) examined the GDP, energy and ecosystem of BRICS nations. in these economies. Finally, a recent study on renewable energy
However, the findings confirmed the conservative and feedback consumption was conducted by Su et al. (2021) for 7 OECD coun-
hypothesis for South Africa and Russia; however, the neutrality tries using CCE-MG and CS-ARDL methods. The study findings
hypothesis held for Brazil, India and China. For 6 GCC economies, revealed that R&D in renewable energy consumption, political risk,
Jammazi and Aloui (2015) employed a Wavelet window analysis and fiscal decentralization enhance renewable energy
and revealed the bidirectional causal relationship between GDP consumption.
and energy use. Further, the study provided the evidence of one-
way causality from CO2 to GDP. Another study by Charfeddine 2.3. Renewable energy at disaggregate level with GDP and EF
(2017) used data from 1997 to 2015 and applied Markov switching
equilibrium correction approach to evaluate the association among Currently, the number of studies regarding renewable energy in
economic development, EF, energy use and CO2. The results the investigation of GDP and ecosystems is increasing. Still, only a
depicted that economic development and energy consumption few authors researched renewable energy at a disaggregate level
enhanced the EF but reduced Qatar’s CO2 emission and carbon (Adewuyi and Awodumi, 2017; Destek and Aslan, 2017; Maji
footprint. et al., 2019; Shah et al., 2020; Magazzino et al., 2020).
Using VECM, Wang et al. (2018) studied 170 countries and For instance, Sari et al. (2008) suggested that employment and
explored the significant association among urbanization, GDP, output are forcing indicators for all resources of energy consump-
energy use and CO2. Raza et al. (2019) used Wavelet analysis and tion at a disaggregated level in the long run. Findings by Ohler and
found a positive influence of CO2 discharge on energy consump- Fetters (2014) revealed the most significant positive influence of
tion. Dabachi et al. (2020) studied OPEC African countries and uti- hydroelectric, biomass and waste electricity production on GDP.
lized Larsson’s Heterogeneous Panel Co-integration approach. The A different study by Ahmed et al. (2016) explored that technolog-
results revealed the two-way causal association between energy ical innovation is crucial in reducing CO2, whereas biomass energy
use and GDP, energy prices and GDP, and CO2 and GDP. Finally, has not significant influence on CO2 emissions. For G-7 economies,
from the perspective of Pakistan and Kenya, Baz et al. (2020) exam- Bildirici and Gökmenoğlu (2017) provided evidence of a two-way
ined the association between economic growth, environment and causal link between CO2 and hydropower energy use. Also, unidi-
energy. rectional causality was found from hydropower energy toward
economic growth.
2.2. Studies on renewable energy, GDP and EF Using FMOLS and DOLS techniques, Ahmad and Du (2017)
investigated a study in Iran and quantified a positive influence of
The relationship between renewable energy, economic growth CO2 and energy production from oil on economic growth. Dogan
and CO2 emission has been well studied in previous literature for and Inglesi-Lotz (2017) examined a study in 22 biomass energy-
single-country and multi-country datasets. Bento and Moutinho consuming countries and indicated biomass energy consumption
(2016) examined a study on Italy’s renewable and non- reduced CO2 and supported the EKC hypothesis. Likewise,
renewable energy production. Using the ARDL model, they esti- Adewuyi and Awodumi (2017) used the 3SLS techniques and used
mated that trade enhances the CO2 while renewable energy pro- data from 12 West African economies. They confirmed the feed-
duction lessens the CO2. Likewise, Amri (2017) explored the back hypothesis among CO2, EG and biomass energy in Togo, Mali,
linkage among renewable energy, GDP and CO2 for 72 developing Gambia, Nigeria, Faso and Burkina. Using the panel VAR test,
and developed economies. For empirical analysis, the author Antonakakis et al. (2017) presented the heterogeneous effects of
employed Arellano–Bond and Sargan test using data from 1990 energy resources on EG and CO2 in all the panels.
29
Another study by Destek and Aslan (2017) analyzed the influ- All the variables were transformed into natural logarithms, and
ence of hydropower, wind, biomass and solar energy consumption so equation (1) is written as follows;
on CO2 using the AMG approach and panel bootstrap causality test.
Results of the study showed that biomass, hydroelectricity and LNEF it ¼ ait þ b1 LNRGDPit þ b2 LNRT it þ b3 LNFDIit þ b4 LNURBit
wind energy reduce CO2 while solar energy has an insignificant þ b5 LNEIit þ b6 LNWSit þ eit ð2Þ
influence on CO2. Also, the study considered hydropower energy
as the most influential variable in lessening CO2. In addition, Where ait represents the intercept term, ln denotes the natural
Shah et al. (2020) study used the CCE-MG method and highlighted logarithm, b1 - b6 indicate the elasticity of variables whereas eit
that biomass energy plays a crucial role in economic development. represents the error term.
Moreover, the study confirmed the N-shaped linkage between CO2
and economic growth. Focusing on the top 10 economies with the 3.2. Estimation methods
highest installed wind and solar energy capacity, Bulut and
Menegaki (2020) extended the literature by adding solar energy 3.2.1. CSD tests
production with economic growth. Using AMG, D-H test and Since CSD is a critical problem in econometrics in panel data, we
CCE-MG, their findings predicted the neutrality hypothesis for may estimate inconsistent and biased results by ignoring CSD
solar energy production. It must be highlighted that these studies, (Bilgili and Ulucak, 2018; Grossman and Krueger, 1995; Pesaran,
as mentioned earlier, did not comprise disaggregate renewable 2004). So, the present study tests the CSD using the semi-
energy on the production side. parametric Friedman (1937) CSD test, Pesaran (2004) CSD test
Using the machine learning approach, this gap has been occu- and Bias-adjusted LM test before testing the stationary properties
pied by Magazzino et al. (2020) by focusing on three countries of all the variables in all the regions.
India, China and the USA. Results predicted that wind and solar Since we have fixed T and large N, using the Pesaran CSD test,
energy production caused a reduction in CO2 in sampled nations Friedman test, and Bias-adjusted LM test are appropriate. Pesaran
except for India. Following the empirical literature, we perceive CSD test is valid for fixed T with large N and may be expressed as,
that a plethora of literature is available on renewable energy con- sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi h i
sumption, economic growth and CO2 emission at the aggregate 2T X
N1 X
N b 2ij E ðT K Þ q
ðT k Þ q b 2ij
CD ¼ h i ð3Þ
level. Also, the literature has dealt with renewable energy use with NðN 1Þ i¼1 j¼iþ1 var ðT kÞ q b2 ij
GDP and CO2 at a disaggregate level. Still, literature on renewable
energy from the production side at the aggregate and disaggre- In equation (3), T denotes the time, N indicates the sample size
gated levels is missing. Since renewable energy disaggregates into whereas q2ij represents the pairwise correlation coefficient for each
numerous kinds, such as hydropower, wind and solar, biofuels and
cross-section i.
waste energy etc., this discloses the need for this study to fill the
In heterogeneous panel data, slope homogeneity may cause
research gap. Appendix I presents the literature reviewed on
inconsistent and misleading results (Breitung, 2005). To test the
energy, environment, and GDP.
slope homogeneity phenomenon, Pesaran and Yamagata (2008)
established the method of Swamy (1970) as we have large N and
3. Empirical methodology fixed T, so we employ the slope homogeneity test presented by
Pesaran and Yamagata (2008). Since CSD is found across the panel
3.1. Data and empirical model data cross-sections, first-generation panel unit root tests are not
valid because these tests do not allow CSD in the panel data.
For this study, 43 countries are selected based on the accessibil- We employed the testing procedure developed by Pesaran
ity of data from 1990 to 2017. Appendix II presents the list of these (2007), i.e., cross-sectional augmented dickey fuller test (CADF),
selected countries. For the empirical analysis, we used annual data which allows CSD in the panel data to overcome the limitation of
for all the variables. The data for real GDP, real trade, energy inten- first-generation unit root tests. Conventional cointegration tests,
sity, urbanization and FDI was collected from WDI. In addition, the e.g., Pedroni (2001) test, are not based on CSD, so that they may
data for EF, wind and solar energy production was collected from create biased estimates. Since our data is cross-sectionally depen-
Knoema and IEA. The description of variables is presented in dent, we employed Westerlund panel cointegration, based on
Appendix III. error-correction panel cointegration, to examine the cointegration
Table 1 presents the summary statistics of the main variables relationship among variables (Westerlund, 2007). An empirical
from 1204 observations for the overall panel group. It is clear that description of these tests is given in Appendix IV.
all the variables have an average value with a standard deviation of
4.99 ± 2.85, 21.87 ± 2.13, 0.35 ± 1.19, 1.25 ± 0.60, 4.18 ± 0.27, 3.2.2. Panel long-run methods
0.54 ± 1.50, 1.56 ± 0.35, respectively. In addition, all the variables After confirming the panel cointegration among the variables,
exhibit negative skewness except lnRGDP and lnEI; lnRGDP has we employed CCE-MG, CS-DL and CS-ARDL for the long-run find-
the highest positive skewness. Fig. 1 and Fig. 2 present the vari- ings. According to Pesaran and Smith (1995), Sarafidis and
ables’ matrix plot and correlation results. Robertson (2009), Phillips and Sul (2003), and Paramati et al.
To evaluate the effect of wind and solar energy production and (2017), conventional panel regression methods could be inconsis-
economic growth on EF, we specified the equation (1), which is tent and biased in the presence of CSD.
written as follows; First, the CCE-MG regression method is applied, that is robust to
slope homogeneity. CSD was developed by Pesaran (2006), and it
EF it ¼ f ðRGDPit ; RT it ; FDIit ; URBit ; EIit ; WSit Þ ð1Þ contains the means of dependent and independent variables with
In equation (1), the dependent variable is EF, and all others are their unobserved common properties. It can be written as,
independent variables. Here RGDP represents real GDP; RT is real
yit ¼ ai þ bi xit þ ci yit þ di xit þ ci f t þ eit ð4Þ
trade openness; FDI indicates foreign direct investment; URB is
urbanization; EI denotes energy intensity; and WS indicates wind In equation (4), bi denotes the specific coefficient of the panel,
and solar energy production. The notations i and t denote the yit xit represents the variables, f t indicates the unobserved common
country and year correspondingly. with heterogeneous factors; eit and ai denote the error term and
30
Table 1
Summary statistics.
lnEF lnRGDP lnRT lnURB lnFDI lnEI lnWS

Mean 1.252 21.872 0.359 4.184 0.545 1.560 4.993
Median 1.330 21.943 0.127 4.284 0.768 1.511 5.250
Maximum 2.342 33.900 11.959 4.584 4.461 3.053 11.156
Minimum 0.254 14.900 1.716 3.240 10.571 0.666 0.000
Std. Dev. 0.606 2.133 1.193 0.273 1.501 0.353 2.850
Skewness 0.326 0.219 3.759 1.222 1.594 0.608 0.206
Kurtosis 1.999 5.652 23.554 4.278 8.952 3.639 2.080
Jarque-Bera 71.542 362.506 24032.34 382.037 2287.429 94.706 50.989
Probability 0.000 0.000 0.000 0.000 0.000 0.000 0.000
Sum 1507.510 26335.051 432.932 5038.269 656.981 1878.293 6011.950
Sum Sq. Dev. 441.873 5477.811 1713.685 90.085 2713.722 150.676 9776.013
Observations 1204 1204 1204 1204 1204 1204 1204
Fig. 1. Matrix plot of selected variables.
intercept term, respectively. The CCE-MG approach estimates lead us to inconsistent estimates due to lag orders. Furthermore,
unbiased and consistent coefficients of observable variables. This the long-run results of CS-ARDL may be sensitive in assessing the
approach provides robust results in the presence of countless weak influence of outliers for the panel group.
factors, which might be due to global and local shocks throughout To solve these flaws of CS-ARDL, the DCCE CS-DL approach was
the panel group (Moscone and Tosetti, 2010). Most importantly, proposed (Chudik and Pesaran, 2015). This technique depends on
these shocks are caused by a non-stationary process (Kapetanios pooling cross-sections with extensive data and dependency under
et al., 2011). the coefficient of non-homogeneity. The DCCE CS-DL approach’s
The DCCE technique is developed on the principle of the mean main advantage is that it performs well for small sample sizes than
group and pooled mean group presented by Pesaran and Smith DCCE CS-ARDL. In the case of the series’s considerable time (T), CS-
(1995) and Pesaran (1997). By using DCCE, we can make the esti- DL enhances the efficiency of short-run results. However, Chudik
mator more consistent by including more lags of CSD in regression. and Pesaran (2015) indicate that the CS-DL technique might not
Moreover, DCCE has the following advantages over the existing be considered superior but complementary to CS-ARDL; therefore,
techniques in the relevant literature (Chudik and Pesaran, 2015). this study used these both methods.
DCCE approach deals with the problem of CSD by taking logs and
average values of all the cross-sectional units. It considers hetero- 3.2.3. Panel causality test
geneous slopes and assumes a common factor to represent the In examining the relationship among macroeconomic variables,
variables. It can handle a small sample size. This technique can also testing causality among variables is an important step that helps
be applied in the presence of structural breaks and un-balance policymakers to make decisions regarding specific policies.
panel data (Ditzen, 2016). For the long-run estimation of coeffi- Dumitrescu and Hurlin (2012) introduced the D-H test for panel
cients, two methods can be applied, i.e., CS-DL and CD-ARDL and data. The D-H panel test is employed in this empirical study to
the study in hand use both methods. inspect the causal association among variables. The D-H test is
Recently, Bai (2009), Chudik and Pesaran (2015) and Moon and based on the Wald statistic of Granger (1969), which allows CSD.
Weidner (2014) solved the issue of error cross-sectional depen- The D-H causality test is tested based on an average of Wald statis-
dence by developing techniques using ARDL methodology to alter tics. Moreover, the null hypothesis and the alternative hypothesis
unobserved common factors. By doing so, we may obtain long- of the D-H test can be written as follows in equations (5) and (6);
run estimates indirectly, e.g., DCCE CS-ARDL. This technique may
31
Fig. 2. Correlation plot.
H0 : bi ¼ 0 i ¼ 1; 2; ; N ð5Þ
HI : bi ¼ 0 i ¼ 1; 2; N; bi –0 i ¼ N1 þ 1; N1 þ 2; N ð6Þ
Equation (7) describes the regression model of the D-H test as.
XM XM
zit ¼ ai þ m
cz
M1 i i;tm
þ bm y
M1 i i;tk
þ eit ð7Þ
In equation (7), ai and bi are fixed in time dimensions, z and y

are variables observed for N individuals and T periods. Whereas
bm m
i and ci are allowed to vary across cross sections. A flow chart
of methods used in the study is presented in Fig. 3.
4. Empirical findings
This segment presents the outcomes of the empirical analysis.

The findings of slope homogeneity tests, CSD tests, stationary tests,
Westerlund cointegration test, CCE-MG, CS-DL and CS-ARDL, and
D-H test are presented here. Fig. 3. Flow chart diagram.
4.1. Estimates of slope homogeneity and CSD tests

Table 2
Estimates of slope homogeneity tests.
The analysis of the study starts with testing slope homogeneity
and CSD in the panel data (Tables 2 and 3). The findings of the slope Tests Statistics p-value
homogeneity test show that the null hypothesis is rejected at 1 % D 24.759 0.000***
and confirms the absence of slope homogeneity. Further, the CSD Dadj 29.574 0.000***
in the model is tested by using three tests by Pesaran, Friedman, Note:***represents 1 % level of significance.
and Frees. All tests showed that the null hypothesis of CSD is
rejected at 1 %, which confirms the CSD within the model (see
Table 3 for details). ous regression. To determine the integration order, we employed
the CADF unit root test. The estimates of the test are offered in
4.2. Results of stationary tests Table 4. This indicates that apart from real GDP, all the variables
are integrated in order one. The real GDP variable is integrated of
After testing the slope homogeneity and CSD, checking the sta- order I(0). Accordingly, this mixed order of integration allows us
tionarity properties of all the variables is necessary to avoid spuri- to employ an error-correction-based Westerlund panel cointegra-
32
Table 3 tion test to approve the presence of long-run association amongst

Results of Pesaran, Friedman and Frees tests. the variables.
Statistics Pesaran Friedman Frees
16.388 115.014 11.484 4.3. Estimates of panel cointegration
Sig. 0.0000*** 0.000*** 0.000***
The estimates of the Westerlund test are reported in Table 5.
Note:***represents 1 % level of significance.
Findings clearly showed that the null hypothesis of no long-run
cointegration is rejected at a 1 % statistical significance level. It
Table 4
means that the log run relationship exists among EF, wind and
Estimates of CADF test. solar energy, real GDP, real trade, urbanization, FDI, and energy
intensity, and it fulfils the fundamental objective of the study.
Variables At I(0) At I(1)
Moreover, these estimates also enable us to investigate the effect
t-statistics Prob. t-statistics Prob. of these variables on EF.
lnEF 1.839 0.296 4.106 0.000***
lnWS 1.873 0.222 3.100 0.000*** 4.4. Estimates of long-run panel approaches
lnRGDP 2.518 0.000*** – –
lnRT 1.721 0.604 3.754 0.000***
lnURB 1.572 0.898 3.345 0.000*** After confirming the long-run panel cointegration among the
lnFDI 1.741 0.551 2.316 0.000*** variables, we employed CCE-MG, CS-DL and CS-ARDL estimators
lnEI 1.427 0.988 2.909 0.000*** for the long-run estimates of the study. The findings of these three
Note:***represents 1 % level of significance. estimators are presented in Table 6. Findings provide us with sim-
ilar and dissimilar estimates for the study’s variables. We begin
with a discussion of estimates of DCCE CS-DL and investigate the
Table 5
effect of real GDP, real trade, urbanization, wind and solar energy,
Results of Westerlund test. FDI, and energy intensity on EF. Results of DCCE CS-DL show that
the variable lnWS has a negative but significant coefficient at
Statistics Value Z-value P-value Robust P-value
10 %, which indicates that wind and solar energy production neg-
Gt 3.738 6.154 0.000*** 0.000*** atively influences EF. In other words, the higher WS production
Ga 8.838 6.067 1.000 0.000***
Pt –22.979 5.652 0.000*** 0.000***
lowers the EF level. However, the variable lnRGDP has a positive
Pa 10.234 2.246 0.988 0.000*** and significant impact on EF.
Concerning the effect of variable lnRT on EF, the finding cleared
Note:***represents 1 % level of significance.
that real trade significantly negatively impacts the EF. The result
indicates that real trade may play a role in reducing the level of
EF. Likewise, the variable lnURB has a significantly negative coeffi-
Table 6
Results of CCE-MG, CS-DL and CS-ARDL estimators.
Variables Coefficient. Standard. Error. z-value Sig. [95 % Confidence. Interval]

CCE-MG estimator
LNWS 0.0217 0.014 1.49 0.135 0.050 0.006
LNRGDP 0.1813 0.063 2.87 0.004*** 0.057 0.305
LNRT 0.184 0.041 4.46 0.000*** 0.265 –0.103
LNURB 9.765 6.450 1.51 0.130 –22.408 2.876
LNFDI 0.004 0.003 1.22 0.221 0.002 0.011
LNEI 0.198 0.077 2.55 0.011*** 0.046 0.350
CS-DL estimator
LNWS 0.0318 0.017 1.84 0.066* 0.065 0.002
LNRGDP 0.181 0.059 3.07 0.002*** 0.065 0.297
LNRT 0.197 0.042 4.66 0.000*** 0.280 –0.114
LNURB 11.748 6.687 1.76 0.079* 24.855 1.357
LNFDI 0.007 0.003 1.82 0.069* 0.0005 0.015
LNEI 0.186 0.088 2.11 0.035** 0.0129 0.359
R-squared = 0.47, CD Statistic = -2.02** (0.0433), R-squared MG = 0.87, Root MSE = 0.04
CS-ARDL estimator
Long run results
LNWS 0.029 0.016 1.76 0.079* 0.061 0.003
LNRGDP 0.190 0.068 2.80 0.005*** 0.057 0.324
LNRT 0.243 0.069 3.50 0.000*** 0.379 –0.106
LNURB 7.195 6.773 1.06 0.288 20.472 6.081
LNFDI 0.005 0.003 1.63 0.102 0.001 0.011
LNEI 0.305 0.123 2.48 0.013*** 0.064 0.546
Short run results
LNWS 0.017 0.015 1.15 0.251 0.048 0.012
LNRGDP 0.189 0.063 2.97 0.003*** 0.064 0.315
LNRT 0.203 0.048 4.25 0.000*** 0.298 –0.109
LNURB 8.407 6.820 1.23 0.218 21.774 4.959
LNFDI 0.006 0.003 1.66 0.096* 0.001 0.013
LNEI 0.263 0.088 2.96 0.003*** 0.089 0.438
ECT [-1] 1.071 0.052 20.43 0.000*** 1.173 –0.968
R-squared = 0.45, CD Statistic = 3.92*** (0.0001), R-squared (MG) = 0.86, Root MSE = 0.04
Note: ⁄ = (10 %), ⁄⁄ = (5 %) and ⁄⁄⁄ = (1 %) level of significance. Heterogeneous constant partialled out.
33
Table 7
Estimates of D-H test.
Null Hypothesis: W-Stat. Zbar-Stat. Prob. Comments

lnRT< lnEF 3.633*** 3.798 0.000 EF M RT
lnEF< lnRT 3.562*** 3.608 0.000
lnRGDP < lnEF 4.192 5.289 1.E-07
lnEF < lnRGDP 3.041** 2.218 0.026 EF ? RGDP
lnWS < lnEF 3.470*** 3.363 0.000 WS M EF
lnEF < lnWS 3.352*** 3.047 0.002
lnURB <lnEF 5.166 7.890 3.E-15
lnEF < lnURB 4.193 5.291 1.E-07
lnFDI < lnEF 2.945** 1.962 0.049 EF M FDI
lnEF <lnFDI 3.564*** 3.614 0.000
lnEI< lnEF 3.789 4.214 3.E-05
lnEF< lnEI 3.643*** 3.825 0.000 EF ? EI
lnRGDP< lnWS 4.414 5.881 4.E-09
lnWS <lnRGDP 3.322*** 2.968 0.003 WS ? RGDP
lnRT< lnWS 6.887*** 12.483 0.000
lnWS <lnRT 4.508 6.133 9.E-10 RT ? WS
lnURB < lnWS 11.682*** 25.282 0.000 URB M WS
lnWS < lnURB 6.377*** 11.120 0.000
lnFDI < lnWS 3.062** 2.273 0.023 FDI ? WS
lnWS < lnFDI 3.863 4.411 1.E-05
LNEI < lnWS 3.245*** 2.763 0.005 EI ? WS
lnWS <lnEI 4.811 6.942 4.E-12
lnRT< lnRGDP 4.257 5.462 5.E-08
lnRGDP < lnRT 4.995 7.433 1.E-13
lnURB <lnRGDP 7.372*** 13.777 0.000 URB M RGDP
lnRGDP <lnURB 6.915*** 12.557 0.000
lnFDI < lnRGDP 3.362*** 3.073 0.002 FDI ? RGDP
lnRGDP < lnFDI 4.624 6.441 1.E-10
lnEI < lnRGDP 3.534*** 3.534 0.000 EI ? RDP
lnRGDP< lnEI 3.812 4.275 2.E-05
lnURB < lnRT 5.429*** 8.592 0.000 URB M RT
lnRT <lnURB 8.919*** 17.906 0.000
lnFDI < lnRT 2.253 0.113 0.909
lnRT < lnFDI 3.238*** 2.742 0.006 RT ? FDI
lnEI< lnRT 3.651*** 3.846 0.000 EI ? RT
lnRT < lnEI 4.471 6.033 2.E-09
lnFDI < lnURB 4.551 6.247 4.E-10
lnURB < lnFDI 3.908 4.532 6.E-06
lnEI < lnURB 6.101*** 10.384 0.000 EI M URB
lnURB < lnEI 5.962*** 10.012 0.000
lnEI < lnFDI 3.779 4.186 3.E-05
lnFDI < lnEI 2.241 0.081 0.934
Note: M represents bi-directional causality and ? represents bi-directional causality.
cient, which implies a significant negative association between wind and solar energy production would decrease EF up to
urbanization and EF. Furthermore, the variable lnFDI has a positive 2.9 %.
and significant influence on EF. Keeping other factors constant, a Moreover, the short-run estimates of the DCCE CS-ARDL estima-
1 % increase in FDI may increase EF by 0.07 %. Finally, the variable tor showed that the variables lnFDI, lnEI and lnRGDP have a posi-
lnEI also has a significant positive coefficient value of 5 %, which tive and statistically significant influence on EF. Despite this, the
states that energy intensity positively influences EF. In addition, variable lnRT has a strong, significantly negative impact on EF.
a 1 % increase in energy intensity may lead to a rise in EF and vice The result indicates that real trade has a robust inverse effect on
versa. EF in the short run. More importantly, the variables lnWS and
Table 6 also presents the results of the DCCE CS-ARDL estimator. lnURB have insignificant negative coefficient values, which indi-
Long-run results showed that all the variables significantly affect cate that wind and solar energy production coupled with urbaniza-
EF except FDI and urbanization. The variable lnEI has a significant tion has an insignificant impact on EF in the short run.
favourable influence on EF at 1 %. It means that there is a positive Concerning the findings of the CCE-MG estimator, we cannot
linkage between these two variables. As a result, the higher the report the significant coefficient value of variables lnWS, lnURB
energy intensity is, the higher the EF will be. and lnFDI. We notice that urbanization, wind and solar energy pro-
Meanwhile, regarding the effect of variable lnFDI, the finding duction have negative and insignificant influences on EF. However,
revealed that FDI has an insignificant positive influence on EF. Like- FDI has an insignificant positive impact on EF. Equally important,
wise, the influence of variable lnURB on EF is also harmful and the variable lnRGDP positively and significantly impacts EF. The
insignificant. In addition, concerning variable lnRGDP, the influ- coefficient value reports that a 1 % rise in real GDP enhances EF
ence of real GDP on EF is found to be significantly positive at 1 %. by 18 %. At the same time, the impact of variable lnEI is positive
The finding implies that a 1 % rise in real GDP may rise 0.19 % in and significant at 1 %. As a result, the coefficient value indicates a
EF. Besides, the variable lnRT has a negative but highly significant 19 % rise in EF due to a 1 % rise in energy intensity.
coefficient value at 1 %. Keeping other factors constant, there Further, we notice that the variable lnTR negatively and highly
would be a 24 % decrease in EF due to a 1 % increase in real trade. significantly influences EF. The results clearly showed that real
In particular, the influence of variable lnWS on EF is inverse and trade has a negative influence on EF. As a result, the higher real
significant at 10 %. So, the finding indicates that a 1 % increase in trade would lower the EF.
34
Table 8
Diagnostic tests.
Multicollinearity Heteroscedasticity Serial Correlation Normality Omitted Variable Bias

(Breusch and Pagan, 1979) (Jochmans and Verardi, 2019) (Jarque and Bera, 1987) (Ramsey, 1969)
Variables VIF 1/VIF
lnRGDP 1.46 0.685 0.42 43.000 2.592 0.72
lnWS 1.40 0.712 (0.515) (1.000) (0.273) (0.539)
lnRT 1.24 0.807
lnEI 1.10 0.906
lnFDI 1.08 0.922
lnURB 1.05 0.950
Note: p-values are given in parentheses.
4.5. Estimates of panel causality test its 1 % rise decreases the EF by 19 %, 24 % and 18 % under the CS-
DL, CS-ARDL and CCE-MG specifications, respectively. This result sup-
The estimated findings of the D-H test are presented in Table 7. ports the fact that foreign trade reduces environmental quality. Con-
The outcomes reported some unusual patterns. Our empirical esti- sequently, foreign trade opens the door for the transfer of cleaner
mates showed the bidirectional causality between real trade and technologies across boundaries, which may help in reducing the EF
EF. In the same way, we also found evidence of bidirectional and several relevant studies like Destek et al. (2018), Kais and Sami
causality between wind and solar energy and EF, urbanization (2016), Destek and Sinha (2020) have reported similar results. How-
and wind and solar energy, urbanization and real GDP, urbaniza- ever, this finding opposed the estimate of Charfeddine (2017).
tion and real trade, and urbanization and energy intensity. How- Furthermore, the estimated finding of foreign direct investment
ever, we found that EF does Granger cause to real GDP at a 5 % is insignificant under DCCE CS-ARDL and CCE-MG, except for the
level of statistical significance. The finding indicates that unidirec- DCCE CS-DL method. A 1 % rise in lnFDI increases lnEF by 0.07 %
tional causality exists between EF and real GDP. In addition, our under DCCE CS-DL specification. The finding supported the pollu-
findings confirm the unidirectional causality from EF to FDI, EF to tion haven hypothesis and concluded that FDI reduces the environ-
energy intensity, real trade to wind and solar energy, wind and mental quality in the panel of countries being used in the study.
solar energy to real GDP, and FDI to wind and solar energy. Further, the outcome depicts that FDI in these countries is not
Furthermore, we found evidence of unidirectional causality from eco-friendly and thus opposes the pollution halo hypothesis. The
energy intensity to wind and solar energy, FDI to real GDP, energy result is analogous to the outcomes of Chandran and Tang
intensity to real GDP, real trade to FDI, and from energy intensity (2013), Malik et al. (2020), Sun et al. (2017), and You and Lv
to real trade. There was no evidence of causality between urbaniza- (2018). However, this finding contradicted the studies by
tion and EF, real trade and real GDP, FDI and energy intensity, and Shahbaz et al. (2015) and Solarin and Al-Mulali (2018).
FDI and urbanization. These findings are similar and dissimilar to The coefficient value of energy intensity is positive and signifi-
those of CCE-MG, DCCE CS-DL and DCEE CS-ARDL estimators. cant under all the methods being used in the study. However, the
magnitudes of lnEI under these methods are pretty different. A 1 %
rise in lnEI increases the EF by 18 %, 30 %, and 19.8 % under DCCE
5. Discussion CS-DL, DCCE CS-ARDL, and CCE-MG respectively, and the same is
reported by the findings of Aydin and Turan (2020). Finally, the
To inspect the influence of wind and solar energy and economic variable wind and solar energy have a negative and significant
growth on EF, this study used CCE-MG, DCCE CS-DL, and DCCE CS- influence on EF for CS-DL and CS-ARDL econometric specifications.
ARDL approaches. All the variables are essential factors and signifi- Under CCE-MG, the finding is not statistically significant. It is
cantly affect EF for the whole panel. The variables lnWS, lnRT and argued that a 1 % rise in lnWS reduces EF by 3.1 % and 2.9 % for
lnURB, are negative, whereas lnFDI, lnRGDP, and lnEI positively CS-DL and CS-ARDL methods, respectively. This outcome is reliable
impact EF. The finding represents that the real GDP has a significantly with the results of several existing studies (e.g. Jebli et al., 2020;
positive impact on EF for all the estimators. A 1 % increase in lnRGDP Morawska et al., 2018; Mehmood et al., 2017; Shen et al., 2018).
enhances lnEF by 18 %, 19 %, and 18 % under DCCE CS-DL, DCCE CS- Furthermore, the D-H causality test outcomes indicate that wind
ARDL and CCE-MG approach, respectively. This positive influence on and solar energy are important factors triggering EF changes.
real GDP is reasonable because the high-income level stimulates the Furthermore, diagnostic tests are run to ensure the stability of
consumption of resources, human demands, and better quality of the the model’s output, and the findings are provided in Table 8. These
environment. Therefore, GDP has a positive impact on EF in under- findings revealed that the model is free from multicollinearity,
study economies. Similar findings regarding EF and economic heteroscedasticity, serial correlation, normality, non-normality,
growth’s relationship have been reported in the preceding literature and omitted variable bias.
(Danish and Wang, 2019; Jorgenson and Clark, 2009; Malik et al., Furthermore, without additional investigation, it may not be
2020; Pao et al., 2011; Rudolph and Figge, 2017; Solarin and Al- possible to remark on the robustness of the model. The model is
Mulali, 2018; Uddin et al., 2017).
In addition, the influence of urbanization on EF is also negative
Table 9
under selected econometric specifications. This negative influence
Results of FMOLS to check the robustness.
of urbanization validates the earlier finding of Danish et al. (2019),
who specify that urbanization has a mitigating effect on EF in BRICS Variable Coefficient Std. Error t-Statistic Prob.
countries. Moreover, our finding is consistent with ecological mod- lnWS 0.018 0.009 1.931 0.053*
ernization theory and the results of many authors who indicate the lnRGDP 0.180 0.013 13.684 0.000***
lnRT 0.183 0.021 8.548 0.000***
favourable effect of high urbanization on the environment (see Ref.
lnURB 5.873 1.343 4.373 0.000***
Charfeddine and Mrabet, 2017; Charfeddine and Khediri, 2016). lnFDI 0.013 0.002 6.196 0.000***
However, our finding parallels the estimates by Ahmed et al. lnEI 0.208 0.032 6.366 0.000***
(2020) as it has resulted that foreign trade has an inverse impact
Note: ⁄ = (10 %), ⁄⁄ = (5 %) and ⁄⁄⁄ = (1 %) level of significance.
on EF under all the methods used in the study. In the case of lnRT,
35
evaluated using the FMOLS approach in this quest. The robustness energy sources may be promoted. Furthermore, relying only on
check is performed primarily to determine (a) if the associations’ greener energy sources like hydropower, waste, wind, and solar will
turnaround points are within the sample range and (b) if the signs enable us to attain low-carbon, secure, and sustainable development.
of long-run coefficients are the same as determined by Table 8 The policy should be based on a gradual transition from hazardous,
findings. Table 9 summarizes the results of the robustness check. non-renewable energy sources to address future environmental
According to the empirical data, long-run estimates are robust. concerns.
Second, by implementing environmentally friendly consump-
tion and manufacturing practices, these nations should accelerate
6. Conclusion and policy implications their development rates. The divergent trajectories in economic
development and environmental deterioration indicate that these
Many nations have made carbon neutrality one of their main countries have not yet taken action to address the harmful envi-
goals in the fight against climate change. Over the following dec- ronmental externalities brought by the economic growth of their
ades, clean energy is expected to increase at the fastest-growing economies. Therefore, these countries should consider reforming
rate. Renewable energy development is an essential step toward their economic development policies to lessen the usual trade-off
an energy transition in this setting. Wind, tidal, hydroelectric, between economic growth and environmental deterioration. Addi-
wave, geothermal, solar, and biofuels are a few examples of sus- tionally, these economies must implement pertinent regulations to
tainable energy sources. Since each renewable energy source has aid economic and environmental growth. Greening for economic
drawn more and more attention, there is no fuel cycle in which growth policies is essential in this endeavour, and these countries
they are considered more environmentally friendly. The severe should strive to adapt their current production and consumption
environmental effects of current energy sources might be lowered patterns sustainably.
using wind and solar energy. An increasing number of central gov- Third, the nations should adopt the environmentally sustain-
ernments have recently delegated essential environmental protec- able method of production by applying energy-efficient technolo-
tion duties to subnational governments. gies in the export sector and should switch from non-renewables
The current research examined the impact of wind and solar to energy-efficient and less polluting sources to improve the trade
energy production and economic development on EF in 43 nations balance and emphasize exports and production. If policymakers
worldwide between 1990 and 2017. This research article provided concentrate on the technique impact of trade (changes in technol-
robust outcomes by employing three advanced econometric tech- ogy) rather than the scale effect, it might be advantageous for the
niques, i.e., DCCE CS-DL, DCCE CS-ARDL, and CCE-MG. Results indi- nations economically and ecologically since there would be more
cate that wind and solar energy production and real trade decrease productivity and reduced energy usage.
the EF in the countries in the sample. The results, however, show Finally, lowering the economy’s reliance on traditional energy
that the EF is improving due to GDP, energy intensity, and FDI; sources and increasing the efficiency of energy usage in industries
additionally, wind and solar energy production does not affect EF by governments are priorities for achieving sustainable develop-
when using the CCE-MG technique. On the other hand, over the ment. Promoting renewable energy sources like wind, solar, and
long term, the DCCE CS-DL and DCCE CS-ARDL approaches provide nuclear energy might help lower these nations’ high levels of fossil
us with significant estimates. The empirical results showed that, fuel use. On the other hand, boosting energy efficiency and diver-
for all econometric techniques, EF has a significant positive causal sifying energy usage would lessen the environment’s sensitivity
relationship with GDP and energy intensity. Likewise, using CCE- to energy in the under-study nations. Therefore, lowering the
MG, DCCE CS-DL and DCCE CS-ARDL methods, real trade for these energy intensity of an economy will help make energy resources
nations dramatically lowers EF. sustainable and limit environmental damage.
Furthermore, the results of the DCCE CS-DL technique are sig- The research still has certain limitations, although it offers
nificant regarding FDI and urbanization, whereas the results of intriguing insights into the relationship between wind and solar
CCE-MG and DCCE CS-ARDL methods find insignificant estimates energy production, economic growth, and environmental degrada-
of these variables. Finally, it is evident from the study’s outcomes tion. We only considered 43 nations for our investigation since the
that DCCE CS-DL offers us more reliable estimations than DCCE data wasn’t available. For a more thorough analysis, future
CS-ARDL and CCE-MG specifications. We may therefore conclude research may include additional countries. Future research may
that although GDP increases the EF and wind and solar energy pro- take environment variable proxies other than ecological footprints,
duction helps combat environmental degradation. such as nitrogen and sulphur emissions. Additionally, future stud-
The empirical findings provide a few fascinating policy insights. ies may employ additional renewable energy sources, including
First, even though it is more effective to minimize environmental hydro, biofuels, and waste energy production.
pollution, our empirical study shows that wind and solar energy
production is a significant strategy for alleviating ecological foot- Funding
print. This finding is consistent with current policy recommenda-
tions by the Sustainable Development Goals (2015) and Paris No funding was received for the research.
Agreement (2016), which call for removing barriers to expanding
sustainable energy sources like wind and solar energy. As a result, CRediT authorship contribution statement
selected nations have to give top priority to new wind and solar
energy projects. The government must minimize the production Bilal Hussain: Conceptualization, Methodology. Syed Asif Ali
of fossil fuels and choose renewable energy sources, such as wind Naqvi: Conceptualization, Methodology. Sofia Anwar: Writing –
and solar energy, which are popular forms of energy. original draft, Writing – review & editing. Muhammad Usman:
Along with focusing on households, the governments should tar- Writing – original draft, Writing – review & editing.
get the industrial sector with strong policy guidelines and incentives
to switch to solar energy from any other non-renewable source. In Declaration of Competing Interest
addition, wind and solar energy sources can become more prominent
for electricity needs despite increased expenditures on conventional The authors declare that they have no known competing finan-
energy sources. Since non-renewable energy sources predominate cial interests or personal relationships that could have appeared
and have a limited role in the wind and solar energy, low-carbon to influence the work reported in this paper.
36
Acknowledgements Appendix I
The authors are very grateful to all those who provided help Literature reviewed on energy, environment and GDP.
during this research.
Sr. Author (s) Country (s) Time period Main variables Techniques Findings
no.
1. Sari et al. America 2001–2005 Fossil fuel, ARDL Results suggested that
(2008) (monthly data) employment, employment and output are
hydroelectric, wind, forcing indicators for all
wood, waste, natural resources of energy
gas, coal, solar consumption at disaggregate
energy, industrial level in the long run.
production
2. Apergis and 25 OECD 1980–2005 Real EG, GFCF, labor Larsson panel Evidence of long run
Payne (2010) force, coal cointegration and association was confirmed
consumption panel VAR among real EG, labor force,
GFCF and coal consumption.
3. Pao et al. Russia 1990–2007 Output, EC, CO2 Panel EKC hypothesis was not
(2011) cointegration and supported and output has
Granger causality negative influence on
test emissions.
4. Fei et al. China 1985–2007 EC, EG, CO2 DOLS Evidence of cointegration
(2011) was established amongst EG
and EC.
5. Al-mulali 30 Sub 1980–2008 EG, CO2, EC, domestic Pedroni Results revealed that EC have
and Sab Saharan Africa credit, broad money, cointegration test significant positive effects on
(2012) domestic investment and granger CO2, EG and FD.
causality test
6. Govindaraju India, China 1965–2009 CO2, EG, coal Panel Granger Long run relationship was
and Tang consumption, Causality test found among EG, CO2 and
(2013) coal consumption in China
but not in India
7. Cowan et al. BRICS 1990–2010 CO2, EC, EG Panel Granger Conservative and feedback
(2014) causality test hypothesis were confirmed
in South Africa and Russia
respectively. However,
neutrality hypothesis was
hold for other remaining
three countries.
8. Ohler and 20 OECD 1990–2008 GFCF, labor, EG, REP ECM Findings revealed the largest
Fetters by source wise positive influence of
(2014) hydroelectric, biomass and
waste electricity production
on economic growth.
9. Jammazi and Six GCC 1980–2013 EG, CO2, EC Wavelet window Bidirectional causality exists
Aloui (2015) nations analysis between EC and EG, whereas
evidence of unidirectional
causality was found from CO2
to EG.
10. Bento and Italy 1960–2011 CO2, REP, NREP, TO, ARDL Results revealed that TO have
Moutinho EG positive impact on CO2 in
(2016) long run whereas, REP
reduces CO2.
11. Ahmed et al. 24 EU nations 1980- Technological ARDL Technological innovation
(2016) 2010 innovation, EC, CO2, plays a key role in reducing
biomass energy CO2 whereas biomass energy
has no significant influence
on CO2.
(continued on next page)
37
Appendix I (continued)
no.
12. Bildirici and G-7 1961–2013 Hydropower EC, EG, MS- Granger Study provided the evidence
Gokmenoglu economies CO2 causality, MS-VAR of two way causal link
(2017) amongst CO2 and
hydropower energy use. Also,
one way causal link was
found from hydropower
energy toward economic
growth.
13. Ahmad and Iran 1971–2017 GFCF, EP from oil, FMOLS and DOLS CO2 and EP have positive
Du (2017) POP, INF, agriculture influence on EG.
land, CO2, FDI
14. Dogan and 22 biomass 1985–2012 TO, EC, CO2, URB, FMOLS Results indicated that
Inglesi-Lotz consuming biomass energy biomass energy consumption
(2017) nations reduces the CO2 and
supported EKC hypothesis.
15. Adewuyi and 12 West 1980–2010 FD, biomass EC, EG, 3SLS Results confirmed the
Awodumi African TO, URB, physical feedback hypothesis among
(2017) economies capital, CO2, human CO2, EG and biomass energy
capital in Togo, Mali, Gambia,
Nigeria, Faso and Burkina.
16. Antonakakis 106 countries 1971–2011 EC, EG, CO2, gas Panel VAR test Estimates showed the
et al. (2017) by income energy, oil energy, heterogeneous effects of
wise coal energy, EC, REC different sorts of energy
resources on EG and CO2 in
all the panels.
17. Amri (2017) 72 Developing 1990–2012 REC, EG, TO Arellano–Bond Feedback linkage confirmed
and developed and Sargan test between EG and TO, REC and
countries EG as well as between REC
and TO.
18. Charfeddine Qatar 1970–2015 ED, CO2, EF, EC Markov switching FD and EC increase EF
(2017) equilibrium whereas reduces carbon
correction footprint and CO2.
approach
19. Dong et al. 128 countries 1990–2014 EG, POP, CO2, REC CCEMG Results cleared that EG and
(2018) intensity POP positively enhance the
CO2. Further, level of CO2
lowered by RE intensity in all
the regions.
20. Wang et al. 170 countries 1980–2011 EG, URB, EC, CO2 VECM Findings confirmed the
(2018) according to positive relationship among
income levels the variables in long run.
21. Zafar et al. APEC 1990–2015 REC, EG, NREC, TO, CUP-FMOLS TO and RandD expenses have
(2019) GFCF positive influence on EG as
well as EC play a key role in
stimulating EG.
22. Sharif et al. Top 10 1990–2017 REC, CO2 Q-Q regression Negative association was
(2020) polluted and granger found between CO2 and REC.
countries causality in
quintiles
23. Raza et al. USA January 1973-July CO2, EC, EG Wavelet analysis Findings showed that CO2
(2019) 2015 positively influenced in all
periods by EC. Also, evidence
of unidirectional causality
was confirmed from EC to
CO2.
38
no.
24. Destek and G-7 1991–2014 Hydropower, wind, AMG and panel Results showed that biomass,
Aslan (2017) biomass, solar EC, bootstrap hydroelectricity and wind
CO2, EG causality test energy reduces CO2 while
solar energy has insignificant
influence on CO2. Also, study
considered hydropower
energy as a most efficient
energy to reduce CO2.
25. Chen et al. China 1980–2014 CO2, EG, TO, REP, ARDL and VECM Results revealed that EG and
(2019) NREP NREP enhances CO2 whereas
TO and REP reduces CO2.
26. Maji et al. 15 West 1995- Political stability, DOLS Findings showed that using
(2019) African 2014 capital, labor, EG, inefficient and unclean
nations REC, biomass EC energy resources, REC
reduces the productivity as
well as EG.
27. Jebli et al. 102 countries 1990–2015 Industrial value GMM and granger REC has negative influence
(2020) by income added, service value causality test on industrial and service
wise added, CO2, REC value added in the global
panel.
28. Mutascu and EU 1960–2014 CO2, TO Wavelet based TO is accelerated by CO2 in
Sokic (2020) analysis short as well as in long run.
29. Dabachi et al. OPEC African 1970–2018 Energy intensity, EG, Larsson’s Bidirectional causality was
(2020) countries CO2, energy prices, EC Heterogeneous confirmed between, EG and
Panel Co- EC, EG and energy prices, EG
integration and CO2.
30. Adedoyin BRICS 1990–2014 EG, CO2, coal rents ARDL EG and coal rents have
et al. (2020) significant negative influence
on CO2.
31. Baz et al. Pakistan 1971–2014 EG, EF, GFCF, EC Nonlinear ARDL Findings confirmed the
(2020) and asymmetric asymmetric cointegration
causality among the variables.
technique
32. Malik et al. Pakistan 1971–2014 Income, oil price, CO2, ARDL and non- EG and FDI intensify CO2 in
(2020) FDI linear ARDL both periods.
Oil price increases emissions
in short run whereas reduce
CO2 in long run.
33. Destek and OECD 1980–2014 EG, EF, NREC, TO MG, DOLS-MG, U-shaped relationship exists
Sinha (2020) FMOLS-MG and between EG and EF and REC
CCE-MG reduces EF whereas NREC
increases EF.
34. Shahbaz 38 NREC 1990–2018 EG, REC, NREC, GFCF, FMOLS and DOLS REC, NREC, GFCF and labor
et al. (2020) countries labor positively enhance EG.
35. Atil et al. Pakistan 1972–2017 Economic Covariability Findings revealed that oil
(2020) globalization, real EG, analysis prices, natural resources and
real oil price, FD EG have positive influence on
FD.
36. Shah et al. 38 Asian 1990–2017 CO2, ED, biomass EC, CCEMG, panel VAR Results revealed that
(2020) countries INF, URB,GFCF, TO biomass energy play a key
role in ED. Moreover, study
confirmed the N-shaped
linkage between CO2 and EG.
37. Wang et al. BRICS 1990–2016 HD, REC, EG, public Driscoll-Kraay Findings revealed that REC
(2020) countries debt robust standard stimulated HD whereas
errors estimates, public debt has inverse
Westerlund influence on HD.
cointegration and
D-H causality test
(continued on next page)

39
no.
38. Awodumi Top 5 oil 1980–2015 CO2, EG, EG2, URB, TO, Non-linear ARDL Overall results revealed that
and Adewuyi producers in per capita gas and method NREC promotes
(2020) Africa petroleum EC, carbon environmental quality and
intensity EG.
39. Rahman et al. China 1981–2016 EP and EC of coal, CCR model, FMOLS Results confirmed the
(2020) natural gas and crude and VECM based positive influence of
oil, EG granger causality production and consumption
of oil, gas and coal on EG.
40. Rahman and 5 South Asian 1990- EG, REC, NREC, labor, FMOLS, DOLS and The study confirmed the
Velayutham nations 2014 FD, GFCF D-H causality test positive influence of GFCF,
(2020) REC and NREC on EG.
41. Le and Bao 1990–2014 GFCF, institutional MG, AMG and Results showed that EG
(2020) 16 EMDEs quality, EG, REC, CCEMG positively affected through
economies NREC, TO, FD, NREC and REC along with TO,
government FD, GFCF and government
expenditures expenditures.
42. Przychodzen 27 transition 1990–2014 EG, UN, REP, RandD ANOVA and Government debt, EG and UN
and economies expenses, INF, FDI correlation accelerated REP.
Przychodzen inflows, government analysis
(2020) debt, CO2, availability
of domestic credit
43. Bulut and Top 10 1999–2015 EC, GFCF, solar EP, AMG, D-H test, Findings predicted the
Menegaki economies labor CCEMG neutrality hypothesis for
(2020) with highest solar EP.
installed
capacity of
wind and
solar energy
44. Su et al. 7 OECD 1990–2018 Fiscal CCEMG and CS- Results revealed that R&D in
(2021) countries decentralization, R&D ARDL REC, fiscal decentralization,
in REC, economic political risk and economic
innovation, political innovation positively
risk, REC enhance REC.
45. Magazzino India, China, 1983–2017 (USA) EG, CO2, wind and Machine learning Results predicted that wind
et al. (2021) USA 1986–2017(India) solar EP, coal and solar EP caused a
1990–2017 production reduction in CO2 in China and
(China) USA, whereas the findings
were opposite in case of
India.
Note for acronyms: EG = Economic growth; CO2 = Carbon emissions; EF = EF; REC = Renewable energy consumption; NREC = Non-renewable
energy consumption; R&D = Research and development; EKC = Environment Kuznets Hypothesis; ED = Economic development; FD = Financial
development; EC = Energy consumption; EP = Energy production; REP = Renewable energy production; NREP = Non-renewable energy pro-
duction; UN = Unemployment; URB = urbanization; INF = Inflation; GFCF = Gross fixed capital formation; FDI = Foreign direct investment;
TO = Trade openness; HD = Human development; GHG = Greenhouse gases; POP = population; ARDL = Autoregressive distributive lag model;
3SLS = Three stage least square; MS-VAR = Markov switching- vector autoregressive; QARDL = Quintile autoregressive distributive lag model;
Q-Q = Quintile-on-quintile; ANOVA = Analysis of variance; MG = Mean group; AMG = Augmented mean group; CCEMG = Common Correlated
Effects mean group; FMOLS = Fully modified ordinary least square; DOLS = Dynamic ordinary least square; VECM = Vector error correction
model; GMM = Generalized method of moments; PLS = Partial least square; CUP-FMOLS = Continuously updated fully modified OLS; CS-
ARDL = Cross-sectional ARDL; VAR = Vector auto-regression; D-H causality = Dumitrescu and Hurlin causality test; (EMDEs) = Latin America
and Caribbean Emerging Market and Developing Economies; G-7 = Group of seven; BRICS = Brazil, Russia, India, China, South Africa;
APEC = Asia-Pacific Economic Cooperation countries; OECD = Organization for Economic Co-operation and Development; OPEC = Organization
of the Petroleum Exporting Countries; GCC = Gulf Cooperation Council; USA and EU = United states of America and European union
respectively.
40
Appendix II
List of Countries.
United States Canada Mexico Jordan Dominican Republic Germany
Netherlands Spain Italy Turkey United Kingdom France

Belgium Romania Finland Israel North Macedonia Iran
Sweden Norway Greece Peru Nicaragua China

Austria Hungary Switzerland Brazil El Salvador Ecuador
Thailand Australia New Zealand Philippines Morocco Egypt
Guatemala Ireland Costa Rica India Japan Indonesia
Russia
Appendix III
Description of variables.
Variables Signs Definition (Measurement) Data

Source
Ecological footprint lnEF Natural log of ecological footprint Knoema
(GHA per person)
Real foreign trade lnRT Natural log of Real Trade (merchandized exports + merchandise imports / GDP and WDI
then divided by consumer price index)
Real GDP lnRGDP Natural log of Real GDP (in current US$) (GDP/consumer price index) WDI
Urbanization lnURB Natural log of urbanization (% of total population) WDI
Foreign Direct lnFDI Natural log of FDI, net inflows (% of GDP) WDI
Investment
Energy intensity lnEI Natural log of energy intensity (MJ) WDI
Wind and Solar energy lnWS Natural log of Wind and Solar energy production (metric tons of oil equivalent) IEA
production
Appendix IV Panel unit root test
Slope homogeneity test Cross-sectional IPS test (CIPS) and cross-sectional augmented
dickey fuller test (CADF) allows CSD in the panel data to overcome
Pesaran and Yamagata (2008) is based on the homoscedasticity the limitation of first-generation unit root tests (Pesaran, 2007).
assumption framework for fixed N relative to T as presented in The CADF can be calculated as follows;
equations (8), (9) and (10),

i ! Dyi;t ¼ ai þ bi yi;t1 þ ci yt1 þ di Dyi;t þ eit ð11Þ
pffiffiffiffi N1 s k
D¼ N pffiffiffiffiffiffi ð8Þ
In the equation (11), yi;t and yt1 represents the first difference
2k
of individual series and cross sectional averages of lagged level
respectively.
0 1
i
pffiffiffiffiB N1 s k C
Dadj ¼ N@qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiA ð9Þ 1X N
2kðTk1Þ yt1 ¼ y ð12Þ
Tþ1 N i¼1 i;t1
X
N 0
xi M s xi
1 X
0 N
S¼ ðbi bWFE Þ ðbi bWFE Þ ð10Þ
r2i Dyi;t ¼ Dyi;t ð13Þ
i¼1 N i¼1
In the above equations bi represents the coefficients of pooled
OLS regression for each individual i that are ranging from 1 to N.
WhereasM s , K and r2i identifies the identity matrix, number of 1X N
CIPS ¼ CADF i ð14Þ

independent variables and estimates ofr2i . Furthermore, Dadj and N i¼1

D denotes the biased adjusted statistics and standardized We can compute CADF statistic by averaging the CADF i and
dispersion. CADF i in equation indicates the t-statistics.
41
Panel cointegration test Bento, J.P.C., Moutinho, V., 2016. CO2 emissions, non-renewable and renewable
electricity production, economic growth, and international trade in Italy.
Renew. Sustain. Energy Rev. 55, 142–155.
Westerlund (2007) cointegration tests can be expressed as; Bertram, C., Luderer, G., Pietzcker, R.C., Schmid, E., Kriegler, E., Edenhofer, O., 2015.
Complementing carbon prices with technology policies to keep climate targets
0
0
Xk X
k within reach. Nat. Clim. Chang. 5 (3), 235–239.
DY i;t ¼ dt dt þ qi Y i;t1 bi X i;t1 þ /ij Y i;tj þ /ij X i;tj þ li;t Bhattacharya, M., Churchill, S.A., Paramati, S.R., 2017. The dynamic impact of
j¼1 j¼1 renewable energy and institutions on economic output and CO2 emissions
across regions. Renew. Energy 111, 157–167.
ð15Þ Bildirici, M.E., Gökmenoğlu, S.M., 2017. Environmental pollution, hydropower
energy consumption and economic growth: Evidence from G7 countries.
In equation (15), qi represents the speed of adjustment of the Renew. Sustain. Energy Rev. 75, 68–85.
system. Bilgili, F., Ulucak, R., 2018. Is there deterministic, stochastic, and/or club
convergence in ecological footprint indicator among G20 countries? Environ.
Westerlund (2007) cointegration test is based on the least Sci. Pollut. Res. 25 (35), 35404–35419.
squares results of qi that have group mean statistic of Westerlund Blindheim, B., 2015. A missing link? The case of Norway and Sweden: Does
test is calculated as follows; increased renewable energy production impact domestic greenhouse gas
emissions? Energy Policy 77, 207–215.
1X N Breitung, J., 2005. A parametric approach to the estimation of cointegration vectors

qi
Gs ¼ ð16Þ in panel data. Econ. Rev. 24 (2), 151–173.
^ iÞ
N i¼1 Seðq Breusch, T.S., Pagan, A.R., 1979. A simple test for heteroscedasticity and random
coefficient variation. Econometrica, 1287–1294.
Bulut, U., Menegaki, A., 2020. Solar energy-economic growth nexus in top 10
1 XN countries with the highest installed capacity. Energy Sources Part B 15 (5), 297–
T qi
Ga ¼ 0 ð17Þ 310.
N i¼1 qi ð1Þ 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.
Furthermore, the panel statistics are obtained from the follow- Renew. Sustain. Energy Rev. 24, 445–453.
Charfeddine, L., 2017. The impact of energy consumption and economic
ing formulas: development on ecological footprint and CO2 emissions: evidence from a
Markov switching equilibrium correction model. Energy Econ. 65, 355–374.
b
q Charfeddine, L., Khediri, K.B., 2016. Financial development and environmental
Ps ¼ i ð18Þ
bi
Se q quality in UAE: Cointegration with structural breaks. Renew. Sustain. Energy
Rev. 55, 1322–1335.
Charfeddine, L., Mrabet, Z., 2017. The impact of economic development and social-
b
Pa ¼ T q ð19Þ political factors on ecological footprint: A panel data analysis for 15 MENA
countries. Renew. Sustain. Energy Rev. 76, 138–154.
Chen, Y., Wang, Z., Zhong, Z., 2019. CO2 emissions, economic growth, renewable and
It may be established that cointegration exists in at least one
non-renewable energy production and foreign trade in China. Renew. Energy
cross sectional unit of panel when Gs and Ga reject the null 131, 208–216.
hypothesis. Chudik, A., Pesaran, M.H., 2015. Common correlated effects estimation of
heterogeneous dynamic panel data models with weakly exogenous
regressors. J. econom. 188 (2), 393–420.
References Cowan, W.N., Chang, T., Inglesi-Lotz, R., Gupta, R., 2014. The nexus of electricity
consumption, economic growth and CO2 emissions in the BRICS countries.
Energy Policy 66, 359–368.
Adewuyi, A.O., Awodumi, O.B., 2017. Biomass energy consumption, economic
Dabachi, U.M., Mahmood, S., Ahmad, A.U., Ismail, S., Farouq, I.S., Jakada, A.H., Kabiru,
growth and carbon emissions: fresh evidence from West Africa using a
K., 2020. Energy consumption, energy price, energy intensity environmental
simultaneous equation model. Energy 119, 453–471.
quality, and economic growth nexus in african OPEC countries: evidence from
Ahmad, N., Du, L., 2017. Effects of energy production and CO2 emissions on
simultaneous equations models. J. Environ. Treatment Tech. 8 (1), 403–409.
economic growth in Iran: ARDL approach. Energy 123, 521–537.
Danish, & Wang, Z. (2019). Investigation of the ecological footprint’s driving factors:
Ahmed, A., Uddin, G.S., Sohag, K., 2016. Biomass energy, technological progress and
What we learn from the experience of emerging economies. Sustainable Cities
the environmental Kuznets curve: Evidence from selected European countries.
and Society, 49.
Biomass Bioenergy 90, 202–208.
Danish, Hassan, S. T., Baloch, M. A., Mahmood, N., & Zhang, J. (2019). Linking
Ahmed, Z., Asghar, M.M., Malik, M.N., Nawaz, K., 2020. Moving towards a
economic growth and ecological footprint through human capital and
sustainable environment: the dynamic linkage between natural resources,
biocapacity. Sustainable Cities and Society, 47, 101516.
human capital, urbanization, economic growth, and ecological footprint in
Danish, Zhang, B., Wang, Z., & Wang, B. (2018). Energy production, economic growth
China. Resour. Policy 67, 101677.
and CO 2 emission: evidence from Pakistan. Natural Hazards, 90(1), 27-50.
Ali, R., Kuriqi, A., Abubaker, S., Kisi, O., 2019. Hydrologic alteration at the upper and
Destek, M.A., Aslan, A., 2017. Renewable and non-renewable energy consumption
middle part of the Yangtze river, China: towards sustainable water resource
and economic growth in emerging economies: Evidence from bootstrap panel
management under increasing water exploitation. Sustainability 11 (19), 5176.
causality. Renew. Energy 111, 757–763.
Al-Mulali, U., Sab, C.N.B.C., 2012. The impact of energy consumption and CO2
Destek, M.A., Sinha, A., 2020. Renewable, non-renewable energy consumption,
emission on the economic growth and financial development in the Sub
economic growth, trade openness and ecological footprint: Evidence from
Saharan African countries. Energy 39 (1), 180–186.
organization for economic Co-operation and development countries. J. Clean.
Amri, F., 2017. Intercourse across economic growth, trade and renewable energy
Prod. 242, 118537.
consumption in developing and developed countries. Renew. Sustain. Energy
Destek, M.A., Ulucak, R., Dogan, E., 2018. Analyzing the environmental Kuznets
Rev. 69, 527–534.
curve for the EU countries: the role of ecological footprint. Environ. Sci. Pollut.
Antonakakis, N., Chatziantoniou, I., Filis, G., 2017. Energy consumption, CO2
Res. 25 (29), 29387–29396.
emissions, and economic growth: An ethical dilemma. Renew. Sustain. Energy
Ditzen, J. (2016), Xtdcce: Estimating Dynamic Common Correlated Effects in Stata.
Rev. 68, 808–824.
United States: The Spatial Economics and Econometrics Centre. Online available
Apergis, N., Payne, J.E., 2010. Coal consumption and economic growth: Evidence
at: http://fmwww.bc.edu/RePEc/usug2016/ditzen_uksug16.pdf Accessed on:
from a panel of OECD countries. Energy Policy 38 (3), 1353–1359.
February 2022.
Atil, A., Nawaz, K., Lahiani, A., Roubaud, D., 2020. Are natural resources a blessing or
Dogan, E., Inglesi-Lotz, R., 2017. Analyzing the effects of real income and biomass
a curse for financial development in Pakistan? The importance of oil prices,
energy consumption on carbon dioxide (CO2) emissions: empirical evidence
economic growth and economic globalization. Resour. Policy 67, 101683.
from the panel of biomass-consuming countries. Energy 138, 721–727.
Aydin, M., Turan, Y.E., 2020. The influence of financial openness, trade openness,
Dong, K.Y., Sun, R.J., Li, H., Jiang, H.D., 2017. A review of China’s energy consumption
and energy intensity on ecological footprint: revisiting the environmental
structure and outlook based on a long-range energy alternatives modeling tool.
Kuznets curve hypothesis for BRICS countries. Environ. Sci. Pollut. Res. 27 (34),
Pet. Sci. 14 (1), 214–227.
43233–43245.
Dong, K., Hochman, G., Zhang, Y., Sun, R., Li, H., Liao, H., 2018. CO2 emissions,
Azam, A., Rafiq, M., Shafique, M., Zhang, H., Yuan, J., 2021. Analyzing the effect of
economic and population growth, and renewable energy: Empirical evidence
natural gas, nuclear energy and renewable energy on GDP and carbon
across regions. Energy Econ. 75, 180–192.
emissions: A multi-variate panel data analysis. Energy 219, 119592.
Doytch, N., 2020. The impact of foreign direct investment on the ecological
Bai, J., 2009. Panel data models with interactive fixed effects. Econometrica 77 (4),
footprints of nations. Environ. Sustain. Indicators 8, 100085.
1229–1279.
Dumitrescu, E.I., Hurlin, C., 2012. Testing for Granger non-causality in
Baz, K., Xu, D., Ali, H., Ali, I., Khan, I., Khan, M.M., Cheng, J., 2020. Asymmetric impact
heterogeneous panels. Econ. Model. 29 (4), 1450–1460.
of energy consumption and economic growth on ecological footprint: Using
asymmetric and nonlinear approach. Sci. Total Environ. 718, 137364.
42
Fei, L., Dong, S., Xue, L., Liang, Q., Yang, W., 2011. Energy consumption-economic Mehmood, M.A., Shafiq-ur-Rehman, A.R., Ganie, S.A., 2017. Spatio-temporal changes
growth relationship and carbon dioxide emissions in China. Energy Policy 39 in water quality of Jhelum River. Kashmir Himalaya. Int. J. Environ. Bioener 12
(2), 568–574. (1), 1–29.
Friedman, M., 1937. The use of ranks to avoid the assumption of normality implicit Mele, M., Gurrieri, A.R., Morelli, G., Magazzino, C., 2021. Nature and climate change
in the analysis of variance. J. Am. Stat. Assoc. 32 (200), 675–701. effects on economic growth: an LSTM experiment on renewable energy
Goh, T., Ang, B.W., 2018. Quantifying CO2 emission reductions from renewables and resources. Environ. Sci. Pollut. Res. 28 (30), 41127–41134.
nuclear energy–some paradoxes. Energy Policy 113, 651–662. Mercure, J.F., Pollitt, H., Viñuales, J.E., Edwards, N.R., Holden, P.B., Chewpreecha, U.,
Golpîra, H., Khan, S.A.R., Zhang, Y., 2018. Robust smart energy efficient production Knobloch, F., 2018. Macroeconomic impact of stranded fossil fuel assets. Nat.
planning for a general job-shop manufacturing system under combined Clim. Chang. 8 (7), 588–593.
demand and supply uncertainty in the presence of grid-connected microgrid. Millstein, D., Wiser, R., Bolinger, M., Barbose, G., 2017. The climate and air-quality
J. Clean. Prod. 202, 649–665. benefits of wind and solar power in the United States. Nat. Energy 2 (9), 1–10.
Govindaraju, V.C., Tang, C.F., 2013. The dynamic links between CO2 emissions, Moon, H. R., & Weidner, M. (2014). Dynamic linear panel regression models with
economic growth and coal consumption in China and India. Appl. Energy 104, interactive fixed effects. USC-INET Research Paper, (15-02).
310–318. Morawska, L., Thai, P.K., Liu, X., Asumadu-Sakyi, A., Ayoko, G., Bartonova, A.,
Granger, C.W., 1969. Investigating causal relations by econometric models and Williams, R., 2018. Applications of low-cost sensing technologies for air quality
cross-spectral methods. Econometrica, 424–438. monitoring and exposure assessment: How far have they gone? Environ. Int.
Grossman, G.M., Krueger, A.B., 1995. Economic growth and the environment. Q. J. 116, 286–299.
Econ. 110 (2), 353–377. Moscone, F., Tosetti, E., 2010. Testing for error cross section independence with an
Gullapalli, S., 2020. Unit-6 Global Warming and Climate Change. Indira Gandhi application to US health expenditure. Reg. Sci. Urban Econ. 40 (5), 283–291.
National Open University, New Delhi. Mutascu, M., Sokic, A., 2020. Trade Openness-CO 2 Emissions Nexus: a Wavelet
Hao, Y., Peng, H., Temulun, T., Liu, L.Q., Mao, J., Lu, Z.N., Chen, H., 2018. How harmful Evidence from EU. Environ. Model. Assess., 1–18
is air pollution to economic development? New evidence from PM2. 5 Ohler, A., Fetters, I., 2014. The causal relationship between renewable electricity
concentrations of Chinese cities. J. Clean. Prod. 172, 743–757. generation and GDP growth: A study of energy sources. Energy Econ. 43, 125–
Hdom, H.A., Fuinhas, J.A., 2020. Energy production and trade openness: Assessing 139.
economic growth, CO2 emissions and the applicability of the cointegration Pao, H.T., Yu, H.C., Yang, Y.H., 2011. Modeling the CO2 emissions, energy use, and
analysis. Energ. Strat. Rev. 30, 100488. economic growth in Russia. Energy 36 (8), 5094–5100.
Hu, H., Zhang, X.H., Lin, L.L., 2014. The interactions between China’s economic Paramati, S.R., Mo, D., Gupta, R., 2017. The effects of stock market growth and
growth, energy production and consumption and the related air emissions renewable energy use on CO2 emissions: evidence from G20 countries. Energy
during 2000–2011. Ecol. Ind. 46, 38–51. Econ. 66, 360–371.
Jammazi, R., Aloui, C., 2015. RETRACTED: On the interplay between energy Pedroni, P., 2001. Purchasing power parity tests in cointegrated panels. Rev. Econ.
consumption, economic growth and CO2 emission nexus in the GCC Stat. 83 (4), 727–731.
countries: A comparative analysis through wavelet approaches. Renew. Pesaran, M.H., 1997. The role of economic theory in modelling the long run. Econ. J.
Sustain. Energy Rev. 51, 1737–1751. 107 (440), 178–191.
Jaramillo, P., Muller, N.Z., 2016. Air pollution emissions and damages from energy Pesaran, H.M., 2004. General diagnostic tests for cross-sectional dependence in
production in the US: 2002–2011. Energy Policy 90, 202–211. panels. University of Cambridge, Cambridge Working Papers in Economics, p.
Jarque, C.M., Bera, A.K., 1987. A test for normality of observations and regression 435.
residuals. Int. Stat. Rev./Revue Internationale de Statistique, 163–172. Pesaran, M.H., 2006. Estimation and inference in large heterogeneous panels with a
Jebli, M.B., Farhani, S., Guesmi, K., 2020. Renewable energy, CO2 emissions and multifactor error structure. Econometrica 74 (4), 967–1012.
value added: Empirical evidence from countries with different income levels. Pesaran, M.H., 2007. A simple panel unit root test in the presence of cross-section
Struct. Chang. Econ. Dyn. 53, 402–410. dependence. J. Appl. Economet. 22 (2), 265–312.
Jochmans, K., & Verardi, V. (2019). XTSERIALPM: Stata module to perform a Pesaran, M.H., Smith, R., 1995. Estimating long-run relationships from dynamic
portmanteau test for serial correlation in panel data. heterogeneous panels. J. Econ. 68 (1), 79–113.
Jorgenson, A.K., Clark, B., 2009. The economy, military, and ecologically unequal Pesaran, M.H., Yamagata, T., 2008. Testing slope homogeneity in large panels. J.
exchange relationships in comparative perspective: A panel study of the Econ. 142 (1), 50–93.
ecological footprints of nations, 1975–2000. Soc. Probl. 56 (4), 621–646. Phillips, P.C., Sul, D., 2003. Dynamic panel estimation and homogeneity testing
Kais, S., Sami, H., 2016. An econometric study of the impact of economic growth and under cross section dependence. Econ. J. 6 (1), 217–259.
energy use on carbon emissions: panel data evidence from fifty eight countries. Przychodzen, W., Przychodzen, J., 2020. Determinants of renewable energy
Renew. Sustain. Energy Rev. 59, 1101–1110. production in transition economies: A panel data approach. Energy 191,
Kapetanios, G., Pesaran, M.H., Yamagata, T., 2011. Panels with non-stationary 116583.
multifactor error structures. J. Econ. 160 (2), 326–348. Rahman, M.M., Velayutham, E., 2020. Renewable and non-renewable energy
Khan, S.A.R., Jian, C., Zhang, Y., Golpîra, H., Kumar, A., Sharif, A., 2019. consumption-economic growth nexus: new evidence from South Asia. Renew.
Environmental, social and economic growth indicators spur logistics Energy 147, 399–408.
performance: From the perspective of South Asian Association for Regional Rahman, Z.U., Khattak, S.I., Ahmad, M., Khan, A., 2020. A disaggregated-level
Cooperation countries. J. Clean. Prod. 214, 1011–1023. analysis of the relationship among energy production, energy consumption and
Kuriqi, A., Pinheiro, A.N., Sordo-Ward, A., Garrote, L., 2019. Flow regime aspects in economic growth: Evidence from China. Energy 194, 116836.
determining environmental flows and maximizing energy production at run-of- Ramsey, J.B., 1969. Tests for specification errors in classical linear least-squares
river hydropower plants. Appl. Energy 256, 113980. regression analysis. J. Roy. Stat. Soc.: Ser. B (Methodol.) 31 (2), 350–371.
Le, H.P., Bao, H.H.G., 2020. Renewable and non-renewable energy consumption, Raza, S.A., Shah, N., Sharif, A., 2019. Time frequency relationship between energy
government expenditure, institution quality, financial development, trade consumption, economic growth and environmental quality in the United States:
openness, and sustainable development in Latin America and Caribbean evidence from transportation sector. Energy 173, 706–720.
emerging Market and developing economies. Int. J. Energy Econ. Policy 10 (1), Rudolph, A., Figge, L., 2017. Determinants of ecological footprints: what is the role
242. of globalization? Ecol. Ind. 81, 348–361.
Magazzino, C., 2012. On the relationship between disaggregated energy production Sarafidis, V., Robertson, D., 2009. On the impact of error cross-sectional dependence
and GDP in Italy. Energy Environ. 23 (8), 1191–1207. in short dynamic panel estimation. Econ. J. 12 (1), 62–81.
Magazzino, C., 2017. Renewable energy consumption-economic growth nexus in Sari, R., Ewing, B.T., Soytas, U., 2008. The relationship between disaggregate energy
Italy. Int. J. Energy Econ. Policy 7 (6), 119–127. consumption and industrial production in the United States: An ARDL approach.
Magazzino, C., Mele, M., Schneider, N., 2020. A Machine Learning approach on the Energy Econ. 30 (5), 2302–2313.
relationship among solar and wind energy production, coal consumption, GDP, Shah, S.A.R., Naqvi, S.A.A., Riaz, S., Anwar, S., Abbas, N., 2020. Nexus of biomass
and CO2 emissions. Renew. Energy 151, 829–836. energy, key determinants of economic development and environment: A fresh
Magazzino, C., Mele, M., Morelli, G., 2021. The relationship between renewable evidence from Asia. Renew. Sustain. Energy Rev. 133, 110244.
energy and economic growth in a time of Covid-19: a machine learning Shahbaz, M., Nasreen, S., Abbas, F., Anis, O., 2015. Does foreign direct investment
experiment on the Brazilian economy. Sustainability 13 (3), 1285. impede environmental quality in high-, middle-, and low-income countries?
Magazzino, C., Toma, P., Fusco, G., Valente, D., Petrosillo, I., 2022. Renewable energy Energy Econ. 51, 275–287.
consumption, environmental degradation and economic growth: the greener Shahbaz, M., Raghutla, C., Chittedi, K.R., Jiao, Z., Vo, X.V., 2020. The effect of
the richer?. Ecological Indicators, 139, 108912.Maji, I. K., Sulaiman, C., & Abdul- renewable energy consumption on economic growth: Evidence from the
Rahim, A. S. (2019). Renewable energy consumption and economic growth renewable energy country attractive index. Energy 207, 118162.
nexus: A fresh evidence from West Africa. Energy Rep. 5, 384–392. Sharif, A., Mishra, S., Sinha, A., Jiao, Z., Shahbaz, M., Afshan, S., 2020. The renewable
Maji, I. K., Sulaiman, C., & Abdul-Rahim, A. S. (2019). Renewable energy energy consumption-environmental quality nexus in Top-10 polluted
consumption and economic growth nexus: A fresh evidence from West countries: Fresh insights from quantile-on-quantile regression approach.
Africa. Energy Reports, 5, 384-392. Renew. Energy 150, 670–690.
Malik, M.Y., Latif, K., Khan, Z., Butt, H.D., Hussain, M., Nadeem, M.A., 2020. Shen, Y., Zhou, H., Xu, J., Wang, Y., Zhang, Q., Walsh, T.R., Wang, Y., 2018.
Symmetric and asymmetric impact of oil price, FDI and economic growth on Anthropogenic and environmental factors associated with high incidence of
carbon emission in Pakistan: Evidence from ARDL and non-linear ARDL mcr-1 carriage in humans across China. Nat. Microbiol. 3 (9), 1054–1062.
approach. Sci. Total Environ. 726, 138421. Solarin, S.A., Al-Mulali, U., 2018. Influence of foreign direct investment on indicators
of environmental quality. Environ. Sci. Pollut. Res. 25 (25), 24845–24859.
43
Su, C.W., Umar, M., Khan, Z., 2021. Does fiscal decentralization and eco-innovation Wang, S., Li, Q., Fang, C., Zhou, C., 2016. The relationship between economic growth,
promote renewable energy consumption? Analyzing the role of political risk. energy consumption, and CO2 emissions: Empirical evidence from China. Sci.
Sci. Total Environ. 751, 142220. Total Environ. 542, 360–371.
Swamy, P.A., 1970. Efficient inference in a random coefficient regression model. Wang, Z., Bui, Q., Zhang, B., Pham, T.L.H., 2020. Biomass energy production and its
Econometrica, 311–323. impacts on the ecological footprint: An investigation of the G7 countries. Sci.
Uddin, G.A., Salahuddin, M., Alam, K., Gow, J., 2017. Ecological footprint and real Total Environ. 743, 140741.
income: panel data evidence from the 27 highest emitting countries. Ecol. Ind. Westerlund, J., 2007. Testing for error correction in panel data. Oxf. Bull. Econ. Stat.
77, 166–175. 69 (6), 709–748.
United Nations Industrial Development Organization, 2019. Clean energy access You, W., Lv, Z., 2018. Spillover effects of economic globalization on CO2 emissions: a
for production use. Accessed at: https://www.unido.org/ our- focus/ spatial panel approach. Energy Econ. 73, 248–257.
safeguarding- environment/clean- energy- access- productive- use/ Zafar, M.W., Shahbaz, M., Hou, F., Sinha, A., 2019. From non-renewable to renewable
renewable-energy. energy and its impact on economic growth: the role of research & development
Van, D.T.B., Bao, H.H.G., 2018. The role of globalization on CO2 emission in Vietnam expenditures in Asia-Pacific Economic Cooperation countries. J. Clean. Prod.
incorporating industrialization, urbanization, GDP per capita and energy use. 212, 1166–1178.
Int. J. Energy Econ. Policy 8 (6), 275. Zhang, X.P., Cheng, X.M., 2009. Energy consumption, carbon emissions, and
Wang, S., Li, G., Fang, C., 2018. Urbanization, economic growth, energy economic growth in China. Ecol. Econ. 68 (10), 2706–2712.
consumption, and CO2 emissions: Empirical evidence from countries with
different income levels. Renew. Sustain. Energy Rev. 81, 2144–2159.
44

1 s2.0 S1342937X23000369 Main

Uploaded by

Copyright:

Available Formats

1 s2.0 S1342937X23000369 Main

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

1 s2.0 S1342937X23000369 Main

Uploaded by

Copyright:

Available Formats

Gondwana Research 119 (2023) 27–44

Contents lists available at ScienceDirect

Effect of wind and solar energy production, and economic development

lnEF lnRGDP lnRT lnURB lnFDI lnEI lnWS

Fig. 1. Matrix plot of selected variables.

Fig. 2. Correlation plot.

In equation (7), ai and bi are fixed in time dimensions, z and y

This segment presents the outcomes of the empirical analysis.

4.1. Estimates of slope homogeneity and CSD tests

Table 3 tion test to approve the presence of long-run association amongst

Variables Coefficient. Standard. Error. z-value Sig. [95 % Confidence. Interval]

Null Hypothesis: W-Stat. Zbar-Stat. Prob. Comments

Note: M represents bi-directional causality and ? represents bi-directional causality.

Multicollinearity Heteroscedasticity Serial Correlation Normality Omitted Variable Bias

Note: p-values are given in parentheses.

(continued on next page)

(continued on next page)

United States Canada Mexico Jordan Dominican Republic Germany

Netherlands Spain Italy Turkey United Kingdom France

Sweden Norway Greece Peru Nicaragua China

Variables Signs Definition (Measurement) Data

Appendix IV Panel unit root test

CIPS ¼ CADF i ð14Þ

1X N Breitung, J., 2005. A parametric approach to the estimation of cointegration vectors

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.