







































Global Sustainability Research                                                                       ISSN: 2833-986X                                                 

https://doi.org/10.56556/gssr.v1i2.327 
                                                                  

 

Global Scientific Research    12 
 

Electricity Production Sources and CO2 Emission in OECD countries: Static and Dynamic 
Panel Analysis 
 

Liton Chandra Voumik1, Md. Jamsedul Islam2, Asif Raihan3* 

 

1Department of Economics, Noakhali Science and Technology University, Noakhali 3814, Bangladesh. Email: 

litonvoumik@gmail.com 
2Department of Tourism and Hospitality Management, Noakhali Science and Technology University, 

   Noakhali 3814, Bangladesh. Email: mjirafi.thm@nstu.edu.bd 
3Institute of Climate Change, Universiti Kebangsaan Malaysia, Bangi 43600, Malaysia. asifraihan666@gmail.com 

 

Corresponding Author: Asif Raihan, asifraihan666@gmail.com, ORCID ID: 0000-0001-9757-9730 

Received: 17 October, 2022, Accepted: 08 November, 2022, Published: 09 November, 2022 

 

 

Abstract 

Industrialization, urbanization, population growth, and changes in lifestyle have all contributed to a rise in the OECD countries' 

risk of global warming. The amount of carbon dioxide (CO2) generated from heat and power sources put out is directly related 

to how much electricity they make. Finding out which sources are bad for the environment, and which are not is the primary 

motivation behind this study. The impact of different approaches to energy production on carbon dioxide emissions is analyzed 

using OECD data. The data is analyzed using Quantile Regression (QR), Generalized Method of Moments (GMM), and 

Pooled Ordinary Least Squares (OLS). The study found that CO2 emissions were significantly impacted in a positive direction 

when electricity was generated using coal, oil, or gas. The emissions from coal-fired power plants are the most detrimental. 

The generation of hydroelectricity and other forms of renewable energy can reduce CO2 emissions in all regression models. 

The most compelling evidence of a correlation between CO2 emissions and energy sources was uncovered in this study. In 

order to produce credible findings, the paper used both QR and GMM methods. Important implications for environmental 

policy are drawn from this article's findings. Both are required to lessen our reliance on fossil fuels and promote the 

development of renewable energy sources like solar, wind, and hydroelectricity. 

 

Keywords: Carbon dioxide; Environmental impact; Economics; Renewable; Sustainability 

 

 

 

Introduction 

 

As the global economy grows, CO2 emissions rise because 

of industrialization, deforestation, environmental damage, 

and the world's population (Raihan et al., 2018; Raihan et 

al., 2019; Jaafar et al., 2020; Raihan et al., 2021; Raihan and 

Tuspekova, 2022a). According to the World Meteorological 

Organization (2021), the amount of carbon dioxide in the air 

has increased by 2.5 times since the beginning of the 

industrial revolution. CO2 can be emitted into the 

environment during the generation and consumption of 

energy (Raihan and Tuspekova, 2022b; Raihan et al., 2022a; 

Raihan et al., 2022b). The amount of greenhouse gases 

(GHGs) in the earth's atmosphere has significantly increased 

due to higher energy use and it has had numerous negative 

impacts on the environment, people's health, the economy, 

and many others (Begum et al., 2020; Raihan et al., 2022c; 

Raihan and Tuspekova, 2022c). More than 80% of the 

world's energy needs grew between 2000 and 2010. Most of 

this growth was made possible by using more than twice as 

many fossil fuels. Over 90% of all CO2 emissions and over 

two-thirds of all GHG emissions come from burning fossil 

fuels and biomass (Raihan and Tuspekova, 2022d; Raihan 

et al., 2022d). This makes CO2 the most critical factor in 

determining the overall GHG emissions trend (Raihan and 

Tuspekova, 2022e; Raihan et al., 2022e; Raihan et al., 

2022f). 

Most of the emissions in many countries come from the 

transportation and energy sectors (Raihan and Tuspekova, 

2022f; Raihan and Tuspekova, 2022g). Since 2000, the 

OECD's economy has grown steadily, but CO2 emissions 

related to energy have gone down. This has happened 

because of changes in the economy's structure and the 

energy supply and because manufacturing has become more

https://doi.org/10.56556/gssr.v1i2.327
mailto:litonvoumik@gmail.com
mailto:mjirafi.thm@nstu.edu.bd
mailto:asifraihan666@gmail.com
mailto:asifraihan666@gmail.com


Global Sustainability Research 

Global Scientific Research    13 
 

 energy efficient. Most OECD countries are on track to meet 

their Kyoto Protocol emission reduction goal although the 

progress is not good enough. Because energy use and CO2 

emissions have recently increased, GHG levels will likely 

rise. Approximately 29% of GHG emissions in OECD 

countries come from the energy sector, 24% from 

transportation, 13% from manufacturing, 9% from 

agriculture, 7% from industrial processes, and 3% from 

waste (IEA, 2019). After three years of stability, CO2 

emissions from the world's energy sector hit a record high 

of 32.8 billion tons in 2017. In 2018, the average amount of 

CO2 each person in OECD countries put into the air was 8.7 

tons, while the average amount for the rest of the world was 

4.3 tons.  

Rapid economic expansion in OECD nations has coincided 

with an increase in energy usage for manufacturing, 

industrial, and service-oriented economic activities. Carbon, 

which is mainly used to make electricity and makes up most 

of the area's energy mix, has been the fastest-growing 

energy source in the OECD region over the last ten years. 

Even though this has helped the region grow and become 

more industrialized, it has also increased CO2 emissions 

from energy use and air pollution, which is detrimental to 

human health (Raihan and Tuspekova, 2022h). Access to 

electric power sources has increased dramatically since 

2000. Nonetheless, 45 million individuals living in the 

OECD region still don't have access to electricity, and many 

use biomass energy as their primary source of cooking fuel 

(IEA, 2019). However, energy consumption in OECD 

countries has been rising at a rate of 6% annually, indicating 

that the region's electric grid is having trouble covering its 

costs due to rapid population growth (IEA, 2019). In terms 

of global power consumption, it has been one of the regions 

with the quickest growth rates, with OECD nations leading 

the way. Indeed, the region has come a long way toward its 

objective of ensuring that everyone has access to power by 

2030. 

There may be numerous advantages to producing electricity 

for human usage. It could, however, hurt the environment 

and people's lives (Raihan and Tuspekova, 2022i; Raihan 

and Tuspekova, 2022j). Electricity can be produced in 

several ways, not all of which have the same environmental 

impact. Compared to other forms of renewable energy, such 

as solar power, the environmental impact of coal is 

substantially larger. In addition to nuclear and hydroelectric 

power facilities, oil can also be used to generate electricity. 

Coal, oil, and natural gas are frequently burned to generate 

electricity (Raihan and Tuspekova, 2022k). These factors 

contribute to the increase of carbon dioxide in the 

atmosphere. However, the location of power generation, 

transmission, and distribution can significantly alter the 

environmental consequences of the power system. Electric 

power systems also include transmission networks, 

distribution lines, and power plants that generate electricity 

from various fuel sources. Both the production and 

consumption of these items have the potential to have an 

impact on the natural world. Construction, electricity 

generation, component isolation, and separation fall within 

this category. The functional and construction effects can be 

further subdivided into operational effects (such as fuel 

sources and global and local pollutants) and building effects 

(manufacturing, installation, decommissioning, and 

disposal). However, energy-producing CO2 emissions have 

not been the focus of any GMM or QR-based studies in the 

OECD region. Therefore, this research intends to identify a 

connection between CO2 emissions and the sources of 

power generation in the OECD region using various 

methods, including System GMM, Difference GMM, and 

Quantile Regression. This article illustrates how much 

carbon dioxide is released from coal, gas, and oil electricity 

production in OECD countries. This study also 

demonstrated the advantageous effects of non-conventional 

methods of electricity production on the environment. This 

investigation aims to discover evidence that hydropower 

and other renewable energy sources can aid in decreasing 

waste and pollution. The environmental policy has been 

suggested in the article with the knowledge gained from this 

investigation.  

 

Literature Review 

 

The literature review overviews the relevant publications 

picked for the study endeavor. Numerous works have 

examined the relationship between power generators, 

renewables, and emissions of carbon dioxide. On the other 

hand, it's unclear how things happened because they 

happened in different ways, in different countries, and at 

different times. How much energy we use and how that 

affects CO2 emissions is a big worry for people living now 

and in the future. Global citizens are already planning for 

the future consequences of climate change. Because of the 

harm, it does to the ecosystem; this is a dangerous situation. 

Dantama et al. (2012) reported that electricity influences all 

spheres of society, from the working class to the upper crust. 

According to numbers from 72 different countries, global 

CO2 emissions increased from 67 million metric tons in 

1990 to 134 million metric tons in 2012. So, environmental 

pollution kills more than 150,000 people yearly (Amri, 

2017).  Furthermore, Ozturk (2010) also linked various 

energy sources to economic growth. He examined several 

electricity sources. Population size, the quantity of 

renewable, fossil, and carbon-intensive energy used, 

nuclear-intensive energy used, over dense cities, and 

detrimental air pollutants like PM10, PM2.5, SO2, NO2, 

CO, and benzopyrene have all been related to the rise in 

global CO2 emissions (Talbi, 2017). They also showed how 

these variables enhance global CO2 emissions. Geothermal 

plants emit CO2 due to their high temperatures (300–700 

°F). The association between energy use, GDP expansion, 

and CO2 emissions were examined using Granger causality 

and panel cointegration (Kristmannsdóttir and Armannsson, 

2003). 70 countries' worth of data were analyzed from 1994-



Global Sustainability Research 

Global Scientific Research    14 
 

2013. Granger causality studies show that the two are related 

to one another in terms of energy usage and carbon dioxide 

emissions. Cointegration tests demonstrated a long-term 

relationship between the variables of interest (economic 

growth and energy consumption) and CO2 emissions. As a 

result of process fixes, rising energy use and economic 

growth led to lower CO2 emissions. Economic activity, 

electricity concentration (demand strategy), heat production 

(supply strategy), and carbon emission index are all taken 

into account in this method of categorizing CO2 emissions 

(demand policy effort). EU nations cut CO2 more than non-

EU nations. Reducing thermal power and increasing energy 

efficiency drove policy. These increases may be due to a 

shift in generation mix or higher power use. Scientists 

proved it (Kim et al., 2020).  

Massive energy use and CO2 emissions threaten future 

generations. Droughts, melting glaciers, rising seas, global 

warming, and heat waves are already happening. Negative 

environmental effects threaten the ecology. Between 1992 

and 2018, Awosusi et al. (2022) analyzed the impact of 

BRICS biomass energy use. Environmental degradation is a 

byproduct of economic development, increased use of 

natural resources, and rising levels of gross capital. The 

research of Aydin (2019) relied on BRICS statistics from 

1992-2013. Biomass energy's economic significance was 

emphasized. The heterogeneous panel data study benefited 

several nations. Biomass energy use is key to economic 

growth and reducing imports. Shisong et al. (2018) 

employed a quantile regression panel approach to estimate 

CO2 emissions. According to their analysis, nonrenewable 

energy reduces CO2 emissions the greatest. It is detrimental 

for high-emission nations to have little involvement in 

developing ecologically sound sources of power. Yu et al. 

(2019) investigated the effect of energy output on industrial 

development and long-term economic growth by 

concentrating on the nations with the most significant gains 

in power generation between 2000 and 2018.  

From 1991 to 2018, the BRICS nations' electricity 

production contributed to industrial output and sustained 

economic growth. A few examples of these fields are 

manufacturing, raw materials extraction, mining, and 

chemical synthesis. It has significantly increased the 

electrification of all enterprises and households during the 

previous three decades (Wang et al., 2010). Apparently, the 

causal relationship between energy generation sources and 

CO2 emissions has not been explored in OECD states. In 

addition, no other studies used system GMM, Difference 

GMM, and the QR approach to uncover useful information 

on CO2 emissions from various power production sources in 

OECD countries. Consequently, the present study’s results 

would help close a gap in the existing body of literature by 

using various econometric methods to identify a connection 

between CO2 emissions and the sources of power generation 

in the OECD region.  

 

Methodology 

 

Data and variables of the study 

 

Data for 38 OECD countries were obtained from the World 

Bank's World Development Indicators (WDI) database on 

an annual panel basis from 1986 to 2020. There are currently 

36 member states of the OECD: Slovak Republic, Slovenia, 

Spain, Czech Republic, Denmark, Estonia, Finland, France, 

Netherlands, Norway, Poland, Portugal, Luxembourg, 

Greece, Hungary, Iceland, Ireland, Italy, Latvia, Austria, 

Belgium, Japan, Korea, New Zealand, Israel, Switzerland, 

Canada, Chile, Colombia, Mexico, Costa Rica, the United 

States, Australia, Sweden, Germany, Turkey, the United 

Kingdom, and Lithuania. The following elements are taken 

into consideration (Table 1): 

 

Table 1. A brief overview of the parameters 

Name of the 

Variables 

Variables 

in Log form 

Elaboration of the 

Parameters 

CO2 L(CO2) CO2 emissions (kt) 

Coal L(Coal) 

Percentage of total 

electricity production that 

comes from coal 

Gas L(Gas) 

Percentage of total 

electricity production that 

comes from gas 

Nuclear L(Nuc) 

Percentage of total 

electricity production that 

comes from nuclear 

Hydro L(Hydro) 

Percentage of total 

electricity production that 

comes from hydro 

Oil L(Oil) 

Percentage of total 

electricity production that 

comes from oil 

Renewable L(Renew) 

Percentage of total 

electricity production that 

comes from renewable 

resources excluding hydro 

 

For a complete breakdown of how each variable was 

measured, the median, mode, standard deviation, and 

extreme values are displayed in Table 2. CO2 has an average 

value that is greater than those of the other variables. 

 

 

 

 

 

 



Global Sustainability Research 

Global Scientific Research    15 
 

Table 2. Synopsis of Descriptive Statistics 

Variables Number of observations Mean Standard deviation Minimum Maximum 

L(CO2) 1172 11.17 1.707 5.458 15.57 

L(Coal) 891 2.760 1.787 -6.483 4.587 

L(Gas) 988 2.162 1.756 -4.785 4.599 

L(Oil) 1071 0.864 2.017 -6.091 4.605 

L(Renew) 961 0.467 1.924 -8.028 4.181 

L(Hydro) 968 1.860 2.253 -6.528 4.601 

L(Nuc) 867 3.347 0.789 -2.433 4.477 

 

Econometric model specification 

 

Coal, gas, oil, renewable energy, nuclear, and 

hydroelectricity are the main ways electricity can be 

produced. Though there are many ways to produce 

electricity, but there is no available data for these sources. 

Using the standard methodological approach, this 

investigation controls CO2 emissions by measuring the 

impact of power generation from coal, natural gas, nuclear, 

hydroelectric, oil, and renewable sources. The following 

equation (1) may be used to determine the influence of 

dependent and independent variables. 

 

CO2 = ƒ (Coal, Gas, Oil, Renewable, Hydroelectric, 

Nuclear)      (1) 

 

Notably, no dummy variables were included; all variables 

were classified. This is because it is anticipated that data 

behavior would not change over time. The multivariate 

econometric model is depicted in Equation (2). 

  

(CO2)it = β0 + β1(Coal)it + β2(Gas)it + β3(Oil)it +
β4(Renew)it + β5(Hydro)it + β6(Nuc)it + εit (2) 

 

The log transformation has been taken in Equation (3). 

    

L(CO2)it = β0 + β1L(Coal)it + β2L(Gas)it + β3L(Oil)it +
β4L(Renew)it + β5L(Hydro)it + β6L(Nuc)it + εit 

      

      (3) 

 

where β0 is the intercept term. β1, β2, β3, β4, β5, and β6 are 

the slope coefficients. The ε is present the residual, and i 

presents the cross-section country, t presents the time. 

 

GMM approach 

 

Researchers often utilize both fixed and random effects 

models when studying topics such as international trade, 

environmental policy, and economic development. The 

dynamic GMM model, developed by Arellano and Bond 

(1991) has only recently begun its operations in these 

sectors. This study used models with fixed and random 

effects separately so that their findings can be compared 

with those obtained from our dynamic model. In order to 

conduct an investigation into the CO2 emissions, a two-stage 

"differenced" and "system" GMM model has been included 

in this investigation. Formulae for the differenced GMM 

and system GMM can be displayed as Equation (4) and 

Equation (5). 

 

L(CO2)it = β0 + β1L(CO2)it + β2L(Coal)it +
β3L(Gas)it + β4L(Oil)it + β5L(Renew)it +
β6L(Hydro)it + β7L(Nuc)it + εit   (4) 

 

L(CO2)it − (CO2)it−1 = β0 + [β1L(CO2)it −
(CO2)it−1] + [β2L(Coal)it − (Coal)it−1] + [β3L(Gas)it −
(Gas)it−1] + [β4L(Oil)it − (Oil)it−1] + [β5L(Renew)it −
(Renew)it−1] + [β6L(Hydro)it − (Hydro)it−1] +
[β6L(Nuc)it − (Nuc)it−1] + (ηit − ηit−1) + (εit −  εit−1)
      (5) 

 

Quantile regression (QR regression) 

 

One important use of the quantile regression method is to 

study outcomes that aren't normally distributed and don't 

have linear relationships with predictor factors. Buchinsky 

(1994) states that to describe the possible different effects, 

it needs to identify the qth-quantile (0< q<1) of the 

dependent variable as a temporary distribution, given a set 

of xi variables as follows: 

 

Qq (yit | β0, εit, xit) = β0 + εit
q

+ βi
q

xit  

 (6) 

 

 

where yt indicates the rate of CO2 emission over time.  

As seen by the following objective function, inference from 

the qth quantile regression requires minimizing the absolute 

value of the residual. Quantile regression is presented in 

Equation (7). 

 

 

Qq (βi
q

) = min β ∑ ⃦ n
q,i,t=1 yit − xitβi

q
   ⃦ =

min [ ∑ q│yit − xitβi
q

 │ +i:yit≥xitβ ∑ (1 −i:yit˂xitβ

q)│yit − xitβi
q

 │]   

                                                 (7) 

 



Global Sustainability Research 

Global Scientific Research    16 
 

Results and Discussion 

 

Correlation analysis is crucial to know if there is a positive 

or negative correlation between the variables in the study. 

When two variables have a positive correlation, it can say 

that they are positively covariant. Negative covariance is 

formed when two variables have an inverse relationship. 

Table 3 correlations show a positive relationship between 

L(CO2) and L(Coal), L(Oil), and L(Gas). The correlation 

between L(CO2) and L(Coal) is the highest and most 

significant, while the correlation between L(CO2) and 

L(Gas) is the lowest and least significant (0.014). Similarly, 

the strongest positive correlation is found between L(CO2) 

and L(Coal) and the value is 0.360. There are no correlation 

coefficients that are greater than 0.80 between any of the 

variables in this research. As a result, there is no 

multicollinearity and the study's variables are not linked to 

one another. To a lesser extent than L(Nuc), L(Renew), and 

L(Hydro) have a negative correlation with L(CO2). 

 

 

Table 3. Indicators of variable correlation 
Variables L(CO2) L(Coal) L(Gas) L(Oil) L(Renew) L(Hydro) L(Nuc) 

L(CO2) 1.000       

L(Coal) 0.360*** 1.000      

L(Gas) 0.014 0.212*** 1.000     

L(Oil) 0.197*** 0.480*** 0.089*** 1.000    

L(Renew) -0.098*** -0.017 0.303*** -0.260*** 1.000   

L(Hydro) -0.180*** -0.383*** -0.169*** -0.256*** 0.140*** 1.000  

L(Nuc) -0.353*** -0.286*** -0.423*** -0.172*** -0.262*** 0.264*** 1.000 

*** p<0.01, ** p<0.05, * p<0.1 

  

Table 4. Unit Root Test Result 

Variables 

At Level At 1st Difference 

Harris-Tzavalis 
Im-Pesaran-

Shin 

Levin, Lin & 

Chut  
Harris-Tzavalis 

Im-Pesaran-

Shin 

Levin, Lin & 

Chut  

L(CO2) 0.488 0.847 -0.471 -23.45*** -8.765*** -5.613*** 

L(Coal) 1.254 2.294 4.70 -32.44*** -9.213*** -7.29*** 

L(Gas) -0.94 1.167 0.362 -32.10*** -8.956*** -5.15*** 

L(Oil) 1.2456 -0.863 -0.073 -21.25*** -9.33*** -7.88*** 

L(Renew) -0.98 -0.736 -0.684 -31.83*** -9.247*** -7.82*** 

L(Hydro) -1.11 0.617 0.545 -39.52*** -9.769*** -7.72*** 

L(Nuc) -2.18 -1.054 -1.028 -44.82*** -10.75*** -9.687*** 

*** p<0.01, ** p<0.05, * p<0.1

Empirical economists frequently use panel unit root tests, 

although there is still room for debate on how to interpret 

the results. This note elaborates on how the rejection of the 

panel unit root hypothesis can be read as proof that a sizable 

fraction of the units are stationary. For this reason, it is 

suggested that, in the case of a rejection, especially in 

contexts where the temporal dimension of the panel is quite 

large, the test outcome be supplemented with an estimate of 

the fraction of the cross-section units for which the 

individual unit root tests are rejected. The rejection's 

monetary impact can be gauged by how much this 

percentage rises (Pesaran, 2012). The outcomes of a unit 

root test for the dependent and independent variables are 

shown in Table 4 for both the raw data and the first 

difference levels. Specifically, it shows that H(0) has a unit 

root since it is non-stationary, while H(1) does not. It is 

evident that all of the parameters are locked in at H(1). The 

null hypothesis is rejected with a confidence level of 5% if 

there is even a nominal difference between the two 

variables. The dependent and independent variables are 

anticipated to follow a unit root distribution. Here, the 

investigation would reject the null hypothesis if the p-value 

is less than 0.01. 

As shown in Table 5, the log-log model estimates both fixed 

and moving information panels. The coefficient of CO2 

emissions and electricity output from various sources are 

presented in columns 2 and 3 of the model. The coefficient 

represents a percentage change in CO2 emissions due to a 

percentage change in the independent variables. Our model's 

dynamic panel regression is shown in columns 4 and 5. The 

estimate is statistically significant and positive, with 

coefficients of L(CO2) for explaining L(Coal) of  0.108*** 

and 0.109*** in the fixed and random effect models, 

respectively. The differenced GMM model predicts a 

0.0629 percentage point rise in CO2 emissions for a one 

percent increase in L(Coal), while the System GMM model 

predicts a 0.00802 percentage point decrease. Direct sources 

of CO2 emissions are important, but indirect sources 



Global Sustainability Research 

Global Scientific Research    17 
 

including hard coal, natural gas, and non-energy 

consumption also contribute significantly (Huang et al., 

2018). For cases where both L(Coal) and L(Oil) are major 

contributors to increasing CO2 in a given panel research 

region, the second-phase output is determined by the 

dynamic GMM model. L(Renew) exhibits a -0.0101 and -

0.0101 coefficient weighted negative and significant effect 

on L(CO2) in both the fixed and random effect models. For 

the differenced GMM model, a 1% increase in L(Renew) 

leads to a 0.00605% rise in CO2 emissions, while for the 

System GMM model, the growth is only 0.00485%. 

In the same way, the Differenced GMM model predicts a 

0.0417% drop in CO2 emissions for a 1% rise in L(Hydro), 

while the System GMM model predicts a 0.0042% drop. 

Table 5's Haussmann chi-square value of 18.52** favors the 

fixed-effect model over the random-effects model, with a 

significance level of 0.125. The system GMM model only 

differs considerably in sign and direction from the fixed-

effect model for L(Oil). To ensure the internal consistency 

of the system GMM estimate, it is assumed that the error 

term is not serially correlated. Due to the low AR-2 values 

for both models, we cannot conclude that the first-

differenced error at order 2 is serially correlated. Methods 

that can be relied on are required for use in GMM 

estimations. Applying the Sargan and Hansen tests for over-

identifying constraints in estimation allows one to examine 

the general validity of an instrumental variable. The null 

hypothesis assumed that all instruments were reliable 

because they were considered exogenous. This study shows 

that the Hansen test returns probabilities between 0.243 and 

0.415. On the other hand, the Sargan test statistics of the 

present investigation yield probability values of 0.250 and 

0.284. The null hypothesis that the instruments are reliable, 

was accepted for both. 
 

Table 5. Dynamic and Static Panel Regression Result 

Variables Fixed effect Random effect Differenced GMM System GMM 

L(CO2)   
0.716*** 

(0.0629) 

0.980*** 

(0.00802) 

L(Coal) 
0.108*** 

(0.0142) 

0.109*** 

(0.0141) 

0.0779** 

(0.0335) 

0.0130* 

(0.00881) 

L(Gas) 
0.0797*** 

(0.0113) 

0.0801*** 

(0.0112) 

0.0411*** 

(0.0134) 

0.00583** 

(0.00459) 

L(Oil) 
0.0374*** 

(0.0103) 

0.0377*** 

(0.0103) 

0.00896 

(0.00561) 

0.00861 

(0.00614) 

L(Renew) 
-0.0101 

(0.00665) 

-0.0101 

(0.00662) 

-0.00605 

(0.00520) 

-0.00485 

(0.00318) 

L(Hydro) 
-0.0629*** 

(0.0139) 

-0.0623*** 

(0.0138) 

-0.0417* 

(0.0296) 

0.00421* 

(0.00495) 

L(Nuc) 
-0.0358** 

(0.0161) 

-0.0361** 

(0.0160) 

-0.00215 

(0.00542) 

-0.00666 

(0.00853) 

Constant 
12.15*** 

(0.0847) 

12.07*** 

(0.377) 

3.287*** 

(0.727) 

0.207** 

(0.101) 

Hausman test  18.52**  0.004 

AR-1    0.184 

AR-2     

Hansen Test   0.243 0.415 

Sargan Test   0.250 0.284 

Number of instruments   306 334 

Observations 805 805 675 692 

R-squared 0.334    

Number of countries 28 28 25 26 

*** p<0.01, ** p<0.05, * p<0.1 

 

Furthermore, Column 2 of Table 6 displays the results of an 

OLS regression between CO2 emissions and power 

generation from different sources. Colum 3 through Colum 

5 display the regression quantiles. The QR model considers 

all three of these frequencies: Q25, Q50, and Q75. Coal and 

natural gas power plants produce more carbon dioxide as 

they generate more electricity. In the QR models for Q25, 

Q50, and Q75, the L(Coal) coefficients to explain L(CO2) 

are 0.0607, 0.209***, and 0.481***, respectively. There is 

a negative and significant effect on L(CO2) from both 

L(Renew) and L(Hydro), except for Q25. With quantile- 

 

specific L(Renew) values of 0.0235, -1.102*, and -0.0327, 

respectively, a 1% increase in L(Renew) sources generates 

a CO2 emission barrier of 0.024%, 0.10%, and 0.034%. 

There is a statistically significant inverse relationship 

between carbon emissions and the generation of power from 

renewable and hydroelectric sources; this is an important 

factor to consider. Power generation from renewable, oil, 

and hydroelectric sources can thus contribute to improved 

environmental quality by lowering atmospheric carbon 

emissions. In addition, less carbon dioxide is released into 

the air when power generation is more streamlined. 

 



Global Sustainability Research 

Global Scientific Research    18 
 

Table 6. Quantile regression outcomes 
Variables Panel OLS Q25 Q50 Q75 

L(Coal) 
0.258*** 

(0.0486) 

0.0607 

(0.0830) 

0.209*** 

(0.0637) 

0.481*** 

(0.0336) 

L(Gas) 
0.316*** 

(0.0707) 

0.131 

(0.121) 

0.492*** 

(0.0926) 

0.268*** 

(0.0489) 

L(Oil) 
0.346*** 

(0.0611) 

0.541*** 

(0.104) 

0.211*** 

(0.0801) 

0.0846** 

(0.0422) 

L(Renew) 
-0.0171 

(0.0455)  

0.0235 

(0.0776) 

-0.102* 

(0.0595) 

-0.0327 

(0.0314) 

L(Hydro) 
-0.248*** 

(0.0427) 

-0.249 

(0.0729) 

-0.127*** 

(0.0560) 

-0.08*** 

(0.0295) 

L(Nuc) 
-0.136 

(0.0841) 

-0.216 

(0.143) 

-0.0683 

(0.110) 

0.0791 

(0.0581) 

Constant 10.79*** 

(0.373) 

11.24*** 

(0.636) 

10.39*** 

(0.488) 

10.34*** 

(0.258) 

Observations 405 405 405 405 

Number of ids 32    

R-squared 0.742    

*** p<0.01, ** p<0.05, * p<0.1 

 

In most cases, coal, oil, and gas are the primary fuels utilized 

in the generation of electrical power. Degradation of the 

local environment is exacerbated by these countries' use of 

fossil fuels for energy production, particularly coal, oil, and 

natural gas. The present study focused on answering the 

question of what can be done to reduce the carbon dioxide 

emissions of power plants. It is untrue to assume that the 

only way carbon dioxide is released into the atmosphere is 

through the burning of fossil fuels like coal, oil, and gas. Our 

novel approach to this issue is based on a hybrid of the 

conventional least squares method, a variant of the method 

of moments, and the quantile regression technique. No other 

scientific study has ever attempted something like this 

before. Because this study care about the accuracy of both 

the methods and results, this investigation consistently 

mixes the two. Massive increases in CO2 emissions are 

produced when fossil fuels like gas, coal, and oil are burned. 

It is possible that switching to renewable energy sources like 

hydropower might drastically reduce pollution levels 

(Voumik et al., 2022a; Voumik et al., 2022b; Voumik et al., 

2022c) The GMM method proves that burning coal and 

extracting oil are the two leading causes of CO2 emissions. 

CO2 emissions may be reduced, however, if power is 

generated from renewable and hydroelectric sources, as 

shown by the GMM approach. Quantile regression analysis 

of emissions data provides some context, showing that coal 

and natural gas are the two most important contributors to 

air pollution and, by extension, global warming. 

Hydropower and other renewable energy sources have the 

potential to aid in emission reductions, as shown by the 

GMM and Quantile Regression methods. Hydroelectricity 

and other forms of renewable energy are favored by all 

economic models as the most effective means of minimizing 

harmful effects on the environment (Huang et al., 2018; 

Aydin, 2019). Coal-fired energy is much more damaging to 

the environment than other forms of pollution. This study’s 

aims are met as a result as well. It is generally accepted that 

the environmental impact of electricity generated by 

burning coal is more significant than that of power generated 

by burning oil or natural gas. However, hydroelectricity 

production is better for the environment than renewable 

energy sources that don't use hydropower. This paper's 

contribution quantifies how the independent factors affected 

the dependent ones. Potentially significant long-term 

renewable energy sources include wind, solar, geothermal 

heat, and biomass. The majority of OECD countries get their 

power from coal and oil. Coal and gas consumption in the 

OECD as a whole has expanded quickly since 1980; 

however, consumption has declined in some countries like 

New Zealand and Norway, and demand in the United States 

currently surpasses output, which is rising at a high rate. 

 

Conclusions and Policy Implications 

 

This study examined the relationship between energy 

consumption and carbon dioxide emissions in OECD 

economies. Using panel generalized methods of moments 

(GMM) and quantile regression, this research analyzed time 

series data on OECD nations from 1986 to 2020. The 

research empirically assessed the impact of different power 

generation sources on CO2 emissions using quantile 

regression and generalized linear models (GMM). When 

compared to traditional statistical methods like ordinary 

least squares (OLS), generalized linear models (GLM), and 

additive random effects (ARDL), quantile regression, one 

step-difference GMM, and system GMM provide a more 

comprehensive explanation of the overall dependence of 

energy generation from diverse sources on CO2 emissions. 

As a result of their more significant outputs, coal and natural 

gas have more prominent production factors. Production 

rates of renewable energy are inversely proportional to their 

ecological footprint. Because of the positive ecological 

effects, renewable energy sources should be prioritized 

above traditional ones. CO2 emissions and energy 



Global Sustainability Research 

Global Scientific Research    19 
 

generating sources are positively correlated in OECD 

countries. Hydroelectric power generation also can lessen its 

impact on the environment. Thus, renewable energy sources 

are preferable to conventional electricity generation 

methods. Natural gas and coal are two of the most widely 

used energy sources, and all of our models for OECD 

nations indicated a high positive correlation between the two 

in terms of CO2 emissions. Regardless, the oil-based power 

generation didn't give us any direct orders. 

This article provides a clear and consistent picture of the 

current state of renewable energy sources in OECD 

countries. Since most OECD nations have abundant natural 

resources, this article shows they have a great chance to 

meet their overall power demand by using their renewable 

resources, such as solar, wind, geothermal, hydropower, 

ocean energy, and bioenergy. If they can maximize their 

usage of renewable energy sources, not only will their 

electricity needs be met at a low cost, but their CO2 output 

may also be kept to a minimum. So, instead of relying solely 

on conventional ways, now is the time to look forward and 

collaborate with these renewable energy areas to produce 

electricity. There has already been some effort made by 

OECD governments to resolve the power situation. 

Alternative energy sources, such as solar power, should be 

given more attention because they represent a promising 

new industry with the potential to supply the vast majority 

of the OECD's population's energy and power needs. 

Combined with other forms of renewable energy, these can 

significantly help meet their daily power needs. OECD 

countries may be able to meet their future electricity needs 

using these resources to generate electricity. To alleviate the 

power shortage situation in their respective regions, 

governments and the private sectors should collaborate to 

emphasize using renewable energy sources for electricity 

generation. The article also provides a brief overview of the 

current state of renewable energy worldwide. With fossil 

fuel reserves continuously dwindling, the world must 

increasingly rely on renewable energy sources to keep up 

with increasing energy demands. By maximizing the 

utilization of renewable energy sources, this article is the 

first step toward building a completely uninterrupted power 

flow that is both environmentally friendly and efficient. 

 

Funding: Not applicable 

 

Acknowledgments: Not applicable 

 

Conflict of interest: The authors declare no conflict of 

interest. 

 

References 

 

Amri, F. (2017). Intercourse across economic growth, trade 

and renewable energy consumption in developing 

and developed countries. Renewable and 

Sustainable Energy Reviews, 69, 527-534. 

https://doi.org/10.1016/j.rser.2016.11.230 

Arellano, M., & Bond, S. (1991). Some tests of specification 

for panel data: Monte Carlo evidence and an 

application to employment equations. The review of 

economic studies, 58(2), 277-297. 

https://doi.org/10.2307/2297968 

Awosusi, A. A., Adebayo, T. S., Altuntaş, M., Agyekum, E. 

B., Zawbaa, H. M., & Kamel, S. (2022). The 

dynamic impact of biomass and natural resources on 

ecological footprint in BRICS economies: a quantile 

regression evidence. Energy Reports, 8, 1979-1994. 

https://doi.org/10.1016/j.egyr.2022.01.022 

Aydin, M. (2019). The effect of biomass energy 

consumption on economic growth in BRICS 

countries: A country-specific panel data analysis. 

Renewable Energy, 138, 620-627. 

https://doi.org/10.1016/j.renene.2019.02.001 

Begum, R.A., Raihan, A., Said, M.N.M., 2020. Dynamic 

impacts of economic growth and forested area on 

carbon dioxide emissions in Malaysia. 

Sustainability 12(22), 9375. 

https://doi.org/10.3390/su12229375 

Buchinsky, M. (1994). Changes in the US wage structure 

1963-1987: Application of quantile regression. 

Econometrica: Journal of the Econometric Society, 

62, 405-458. 

Dantama, Y. U., Abdullahi, Y. Z., & Inuwa, N. (2012). 

Energy consumption-economic growth nexus in 

Nigeria: An empirical assessment based on ARDL 

bound test approach. European Scientific Journal, 

8(12), 141-157. 

Huang, L., Krigsvoll, G., Johansen, F., Liu, Y., & Zhang, X. 

(2018). Carbon emission of global construction 

sector. Renewable and Sustainable Energy Reviews, 

81, 1906-1916. 

https://doi.org/10.1016/j.rser.2017.06.001 

IEA (2019) International Energy Agency, Southeast Asia 

energy outlook 2019, IEA, Paris. https:// www. iea. 

org/ repor ts/ south eastasia- energy- outlo ok- 2019. 

Accessed 10 Feb 2022 (Accessed 09 August 2022). 

Jaafar, W. S. W. M., Maulud, K. N. A., Kamarulzaman, A. 

M. M., Raihan, A., Sah, S. M., Ahmad, A., Saad, S. 

N. M., Azmi, A. T. M., Syukri, N. K. A. J., Khan, 

W. R. (2020). The Influence of Forest Degradation 

on Land Surface Temperature – A Case Study of 

Perak and Kedah, Malaysia. Forests, 11(6), 670. 

https://doi.org/10.3390/f11060670 

Kim, H., Kim, M., Kim, H., & Park, S. (2020). 

Decomposition analysis of CO2 emission from 

electricity generation: Comparison of OECD 

countries before and after the financial crisis. 

Energies, 13(14), 3522. 

https://doi.org/10.3390/en13143522 

Kristmannsdóttir, H., & Ármannsson, H. (2003). 

Environmental aspects of geothermal energy 

https://doi.org/10.1016/j.rser.2016.11.230
https://doi.org/10.2307/2297968
https://doi.org/10.1016/j.egyr.2022.01.022
https://doi.org/10.1016/j.renene.2019.02.001
https://doi.org/10.3390/su12229375
https://doi.org/10.1016/j.rser.2017.06.001
https://doi.org/10.3390/f11060670
https://doi.org/10.3390/en13143522


Global Sustainability Research 

Global Scientific Research    20 
 

utilization. Geothermics, 32(4-6), 451-461. 

https://doi.org/10.1016/S0375-6505(03)00052-X 

Ozturk, I. (2010). A literature survey on energy–growth 

nexus. Energy policy, 38(1), 340-349. 

https://doi.org/10.1016/j.enpol.2009.09.024 

Pesaran, M. H. (2012). On the interpretation of panel unit 

root tests. Economics Letters, 116(3), 545-546. 

https://doi.org/10.1016/j.econlet.2012.04.049 

Raihan, A., Begum, R. A., Said, M. N. M., Abdullah, S. M. 

S. (2018). Climate change mitigation options in the 

forestry sector of Malaysia. Journal Kejuruteraan 

SI, 1(6), 89-98. http://dx.doi.org/10.17576/jkukm-

2018-si1(6)-11 

Raihan, A., Begum, R. A., Said, M. N. M., Abdullah, S. M. 

S. (2019). A review of emission reduction potential 

and cost savings through forest carbon 

sequestration. Asian Journal of Water, Environment 

and Pollution, 16(3), 1-7. 

https://doi.org/10.3233/AJW190027 

Raihan, A., Begum, R. A., Said, M. N. M., Pereira, J. J. 

(2021). Assessment of Carbon Stock in Forest 

Biomass and Emission Reduction Potential in 

Malaysia. Forests, 12(10), 1294. 

https://doi.org/10.3390/f12101294 

Raihan, A., Begum, R. A., Said, M. N. M., & Pereira, J. J. 

(2022a). Dynamic impacts of energy use, 

agricultural land expansion, and deforestation on 

CO2 emissions in Malaysia. Environmental and 

Ecological Statistics, 29, 477-507.  

Raihan, A., Begum, R. A., Said, M. N. M., & Pereira, J. J. 

(2022b). Relationship between economic growth, 

renewable energy use, technological innovation, 

and carbon emission towards achieving Malaysia’s 

Paris Agreement. Environment Systems and 

Decisions. https://doi.org/10.1007/s10669-022-

09848-0 

Raihan, A., Muhtasim, D. A., Farhana, S., Pavel, M. I., 

Faruk, O., Rahman, M., & Mahmood, A. (2022c). 

Nexus between carbon emissions, economic 

growth, renewable energy use, urbanization, 

industrialization, technological innovation, and 

forest area towards achieving environmental 

sustainability in Bangladesh. Energy and Climate 

Change, 3, 100080. 

https://doi.org/10.1016/j.egycc.2022.100080 

Raihan, A., Muhtasim, D. A., Farhana, S., Hasan, M. A. U., 

Pavel, M. I., Faruk, O., Rahman, M., Mahmood, A. 

(2022d) Nexus between economic growth, energy 

use, urbanization, agricultural productivity, and 

carbon dioxide emissions: New insights from 

Bangladesh. Energy Nexus, 8, 100144. 

https://doi.org/10.1016/j.nexus.2022.100144 

Raihan, A., Muhtasim, D. A., Pavel, M. I., Faruk, O., & 

Rahman, M. (2022e). Dynamic impacts of 

economic growth, renewable energy use, 

urbanization, and tourism on carbon dioxide 

emissions in Argentina. Environmental Processes, 

9, 38. https://doi.org/10.1007/s40710-022-00590-y 

Raihan, A., Muhtasim, D. A., Pavel, M. I., Faruk, O., & 

Rahman, M. (2022f). An econometric analysis of 

the potential emission reduction components in 

Indonesia. Cleaner Production Letters, 3, 100008. 

https://doi.org/10.1016/j.clpl.2022.100008 

Raihan, A., & Tuspekova, A. (2022a). The nexus between 

economic growth, renewable energy use, 

agricultural land expansion, and carbon emissions: 

New insights from Peru. Energy Nexus, 6, 100067. 

https://doi.org/10.1016/j.nexus.2022.100067 

Raihan, A., & Tuspekova, A. (2022b). Toward a sustainable 

environment: Nexus between economic growth, 

renewable energy use, forested area, and carbon 

emissions in Malaysia. Resources, Conservation & 

Recycling Advances, 15, 200096. 

https://doi.org/10.1016/j.rcradv.2022.200096 

Raihan, A., & Tuspekova, A. (2022c). Dynamic impacts of 

economic growth, energy use, urbanization, 

agricultural productivity, and forested area on 

carbon emissions: New insights from Kazakhstan. 

World Development Sustainability, 1, 100019. 
https://doi.org/10.1016/j.wds.2022.100019 

Raihan, A., & Tuspekova, A. (2022d). Role of economic 

growth, renewable energy, and technological 

innovation to achieve environmental sustainability 

in Kazakhstan. Current Research in Environmental 

Sustainability, 4, 100165. 

https://doi.org/10.1016/j.crsust.2022.100165 

Raihan, A., & Tuspekova, A. (2022e). Nexus between 

emission reduction factors and anthropogenic 

carbon emissions in India. Anthropocene Science, 1, 

295-310. https://doi.org/10.1007/s44177-022-

00028-y 

Raihan, A., & Tuspekova, A. (2022f). Nexus between 

economic growth, energy use, agricultural 

productivity, and carbon dioxide emissions: new 

evidence from Nepal. Energy Nexus, 7, 100113. 

https://doi.org/10.1016/j.nexus.2022.100113 

Raihan, A., & Tuspekova, A. (2022g). Dynamic impacts of 

economic growth, energy use, urbanization, 

tourism, agricultural value-added, and forested area 

on carbon dioxide emissions in Brazil. Journal of 

Environmental Studies and Sciences. 

https://doi.org/10.1007/s13412-022-00782-w 

Raihan A, Tuspekova A (2022h) Dynamic impacts of 

economic growth, renewable energy use, 

urbanization, industrialization, tourism, agriculture, 

and forests on carbon emissions in Turkey. Carbon 

Research, 1, 20. https://doi.org/10.1007/s44246-

022-00019-z 

Raihan A, Tuspekova A (2022i) Towards sustainability: 

dynamic nexus between carbon emission and its 

determining factors in Mexico. Energy Nexus, 8, 

https://doi.org/10.1016/S0375-6505(03)00052-X
https://doi.org/10.1016/j.enpol.2009.09.024
https://doi.org/10.1016/j.econlet.2012.04.049
http://dx.doi.org/10.17576/jkukm-2018-si1(6)-11
http://dx.doi.org/10.17576/jkukm-2018-si1(6)-11
https://doi.org/10.3233/AJW190027
https://doi.org/10.3390/f12101294
https://doi.org/10.1007/s10669-022-09848-0
https://doi.org/10.1007/s10669-022-09848-0
https://doi.org/10.1016/j.egycc.2022.100080
https://doi.org/10.1016/j.nexus.2022.100144
https://doi.org/10.1007/s40710-022-00590-y
https://doi.org/10.1016/j.clpl.2022.100008
https://doi.org/10.1016/j.nexus.2022.100067
https://doi.org/10.1016/j.rcradv.2022.200096
https://doi.org/10.1016/j.wds.2022.100019
https://doi.org/10.1016/j.crsust.2022.100165
https://doi.org/10.1007/s44177-022-00028-y
https://doi.org/10.1007/s44177-022-00028-y
https://doi.org/10.1016/j.nexus.2022.100113
https://doi.org/10.1007/s13412-022-00782-w
https://doi.org/10.1007/s44246-022-00019-z
https://doi.org/10.1007/s44246-022-00019-z


Global Sustainability Research 

Global Scientific Research    21 
 

100148. 

https://doi.org/10.1016/j.nexus.2022.100148 

Raihan, A., & Tuspekova, A. (2022j). The nexus between 

economic growth, energy use, urbanization, 

tourism, and carbon dioxide emissions: New 

insights from Singapore. Sustainability Analytics 

and Modeling, 2, 100009. 

https://doi.org/10.1016/j.samod.2022.100009 

Raihan A, Tuspekova A (2022k) Nexus between energy use, 

industrialization, forest area, and carbon dioxide 

emissions: New insights from Russia. Journal of 

Environmental Science and Economics, 1(4), 1-11. 

https://doi.org/10.56556/jescae.v1i4.269 

Shisong, C., Wenji, Z., Hongliang, G., Deyong, H., You, M., 

Wenhui, Z., & Shanshan, L. (2018). Comparison of 

remotely sensed PM2. 5 concentrations between 

developed and developing countries: Results from 

the US, Europe, China, and India. Journal of 

Cleaner Production, 182, 672-681. 

https://doi.org/10.1016/j.jclepro.2018.02.096 

Talbi, B. (2017). CO2 emissions reduction in road transport 

sector in Tunisia. Renewable and Sustainable 

Energy Reviews, 69, 232-238. 

https://doi.org/10.1016/j.rser.2016.11.208 

Voumik, L. C., Hossain, M. S., Islam, M. A., & Rahaman, 

A. (2022a). Power Generation Sources and Carbon 

Dioxide Emissions in BRICS Countries: Static and 

Dynamic Panel Regression. Strategic Planning for 

Energy and the Environment, 401-424. 

https://doi.org/10.13052/spee1048-5236.4143 

Voumik, L. C., Islam, M., Rahaman, A., & Rahman, M. 

(2022b). Emissions of carbon dioxide from 

electricity production in ASEAN countries: GMM 

and quantile regression analysis. SN Business & 

Economics, 2(9), 1-20. 

https://doi.org/10.1007/s43546-022-00318-y 

Voumik, L. C., Rahman, M. H., & Hossain, M. S. (2022). 

Investigating the subsistence of Environmental 

Kuznets Curve in the midst of economic 

development, population, and energy consumption 

in Bangladesh: imminent of ARDL model. Heliyon, 

8(8), e10357. 

https://doi.org/10.1016/j.heliyon.2022.e10357 

Wang, W., Mu, H., Kang, X., Song, R., & Ning, Y. (2010). 

Changes in industrial electricity consumption in 

china from 1998 to 2007. Energy Policy, 38(7), 

3684-3690. 

https://doi.org/10.1016/j.enpol.2010.02.046 

World Meteorological Organization (2021). Report of Sixth 

Methane Reference Gas Inter comparison 

experiment for Asia from 2018 to 2019, World 

Meteorological Organization, GAW Report No. 

264. 

Yu, Z., Liu, W., Chen, L., Eti, S., Dinçer, H., & Yüksel, S. 

(2019). The effects of electricity production on 

industrial development and sustainable economic 

growth: A VAR analysis for BRICS countries. 

Sustainability, 11(21), 5895. 

https://doi.org/10.3390/su11215895 

 

https://doi.org/10.1016/j.nexus.2022.100148
https://doi.org/10.1016/j.samod.2022.100009
https://doi.org/10.56556/jescae.v1i4.269
https://doi.org/10.1016/j.jclepro.2018.02.096
https://doi.org/10.1016/j.rser.2016.11.208
https://doi.org/10.13052/spee1048-5236.4143
https://doi.org/10.1007/s43546-022-00318-y
https://doi.org/10.1016/j.heliyon.2022.e10357
https://doi.org/10.1016/j.enpol.2010.02.046
https://doi.org/10.3390/su11215895

