Rethinking Economic Growth Policies in the Context of Sustainability: Panel Data Analysis on Pollution as an Effect of Economic Development in EU Countries

Abstract

This study investigates the complex interplay between carbon dioxide (CO₂) emissions and significant socio-economic determinants, namely gross domestic product (GDP) per capita, population, and energy consumption. The analysis revealed a deficiency in the literature since most studies have primarily focused on the contemporary period, neglecting the 1970s and 1980s, which were characterized by extensive industrialization in a substantial portion of Europe. The study aims to establish a definitive association between socio-economic factors and the observed fluctuations in CO₂ emissions. The study focuses on a panel of 20 countries within the European Union. It collects 52 yearly observations spanning from 1970 to 2021. The analysis employs panel data regression estimate. Extensive investigation has conclusively demonstrated that a direct and positive correlation exists between population size and energy consumption and the subsequent impact on carbon dioxide (CO₂) emissions. Interestingly, there is a negative association between the GDP per capita and CO₂ emissions, suggesting the possibility of a decoupling between economic growth and carbon emissions. This might be attributed to the development of cleaner technology and a heightened awareness of environmental concerns. Comprehending these processes is crucial for well-informed policy formulation with the objective of attaining a healthy equilibrium between economic success and environmental sustainability. Further investigation is required to explore these associations in greater depth, considering improvements in technology, the efficacy of policies, and variances across different regions, with the aim of promoting a more environmentally friendly and sustainable future.

1. Introduction

Although economic growth is commonly regarded as a favorable measure of a country’s affluence and progress, it can also have adverse consequences on the environment, resulting in escalated levels of pollution [1]. The correlation between economic growth and pollution is attributed to multiple causes and exhibits variability across diverse countries and locations. Industrialization and urbanization are key factors contributing to the positive correlation between economic expansion and pollution [2]. As a nation’s economy expands, there is frequently an increased need for energy, transportation, and manufacturing, all of which have the potential to result in elevated levels of pollution. This encompasses the release of toxins, such as carbon dioxide (CO₂), nitrogen oxides (NO_x), and sulfur dioxide (SO₂), into the atmosphere through emissions from factories, power plants, and cars [3]. Consequently, this contributes to the occurrence of air pollution and the phenomenon of global warming. Moreover, it is frequently observed that as individuals’ financial capacities expand in tandem with economic development, there tends to be a corresponding surge in consumption patterns and a notable transition towards more resource-intensive modes of living [4]. This phenomenon can lead to an increase in both the generation and disposal of garbage, hence increasing environmental issues to a greater extent. Furthermore, the endeavor to achieve economic expansion may occasionally result in the exploitation of natural resources, deforestation, and the destruction of habitats, all of which can have adverse environmental implications [5,6]. Nevertheless, it is crucial to acknowledge that the correlation between economic growth and pollution does not follow a linear pattern. There is a contention that economic expansion has the potential to engender favorable transformations in environmental policies and technologies. As nations experience economic growth, there is a likelihood of increased investment in cleaner technologies and the implementation of more stringent environmental rules, resulting in a decrease in pollution levels [7].

The justification for this study is complex and has multiple aspects. It is crucial to understand the complex relationship between economic growth and environmental deterioration in the European Union (EU), considering its different economies and environmental policies. Through the utilization of panel data analysis, this study can offer an extensive comprehension of this correlation, enabling the formulation of more customized policy suggestions.

The problem statement of the correlation between economic growth and pollution is a diverse and intricate matter that has attracted much academic interest in recent times. The primary objective of the proposed study is to reassess approaches for promoting economic growth while considering sustainability, which is a significant concern in modern economics and environmental policy. There has been an increasing acknowledgment in recent years of the pressing requirement to achieve a harmonious equilibrium between economic progress and ecological sustainability.

The current body of research on the correlation among economic growth, sustainability, and pollution in the EU is lacking a thorough examination that takes into account a longer time period, including the period of forced industrialization in the former communist nations. The lack of research in this area gives a valuable opportunity for the proposed study to enhance the comprehension of the relationship between economic advancement and environmental sustainability in the European Union.

The majority of research in this field generally focuses on the eras after communism or transition, disregarding the significant environmental impacts of the rapid industrialization that occurred in these nations during the communist era. This gap leads to an inadequate comprehension of how historical variables, such as centralized planning and resource-intensive industries, have impacted the long-term environmental performance of these nations.

The proposed study aims to analyze the impact of forced industrialization and its aftermath on present environmental concerns in former communist EU nations. By extending the analysis to cover this historical period, the study will offer valuable insights into the lasting effects of these processes. It will provide insight into the influence of economic policies from that era on current pollution levels and environmental sustainability.

Furthermore, this research will provide essential direction for policy-makers aiming to achieve a harmonious equilibrium between economic expansion and sustainability in nations with diverse historical backgrounds. The study’s novelty comes in its examination of the relationship between economic growth and pollution in EU nations, taking into account their historical context, including experiences with communist industrialization. This addresses a crucial research vacuum.

The main research question of the study is whether the socio-economic and demographic variables influence CO₂ emissions and to what extent. In order to answer to the central question of the research, several research questions were formulated: (1) What is the degree of statistical significance in the link between CO₂ emissions and population in EU nations, and how has this relationship changed over time?; (2) What is the impact of GDP per capita, a measure of economic progress, on CO₂ emissions in the EU, and what are the long-term patterns of this correlation?; and (3) How has energy efficiency and renewable energy adoption altered the connection between EU energy consumption and CO₂ emissions?

This study aims to enhance the ongoing discussion by undertaking a panel data analysis to examine the intricate correlation between economic advancement and pollution in EU nations. Furthermore, a clear correlation between socio-economic factors, such as GDP per capita, population, and energy consumption, and the changes observed in CO₂ emissions will be established. This will be achieved with empirical analysis and econometric testing. In this study, a dataset consisting of 20 nations was utilized, with each country having 50 observations for each variable. The dataset was then subjected to statistical analysis using Eviews 12 statistical software. The methodological approach used in the empirical analysis was based on descriptive statistics, the unit root test, the cointegration test and the estimation of regression.

The primary objective of the proposed study is to reassess approaches for promoting economic growth while considering sustainability, which is a significant concern in modern economics and environmental policy. There has been an increasing acknowledgment in recent years of the pressing requirement to achieve a harmonious equilibrium between economic progress and ecological sustainability. This study aims to enhance the ongoing discussion by undertaking a panel data analysis to examine the intricate correlation between economic advancement and pollution in European Union (EU) nations.

The paper contributes to the existing literature by proving new evidence on a panel of 20 countries member of the European Union by having a broader coverage through the longer timeframe analyzed, between 1970 and 2021, which includes the period of the heavy industrialization in the former communist countries. The study expands the existing empirical research by focusing on the identification of a direct connection between CO₂ emissions and the socio-economic variables representative of achieving economic growth. The research is relevant to the existing literature that examines the effects of economic expansion on environmental deterioration. Furthermore, the study can provide valuable insights into the possible trade-offs and synergies that may exist between economic growth and environmental sustainability. Gaining a comprehensive understanding of these processes is crucial in order to provide guidance to politicians and businesses in developing policies that foster economic development while simultaneously reducing the ecological impact. In summary, this research will provide significant insights to the field of sustainable economics and inform policy decisions based on empirical evidence in the EU and other regions.

The study is organized through six main sections, starting with the current introduction which includes the problem statement and the objectives of the paper followed by the literature review. The analysis phase starts with establishing the research methodology and data from which the empirical results will be extracted. The results are discussed to describe and compare the findings to previous studies.

2. Literature Review

The environmental Kuznets curve (EKC) concept was first introduced by Grossman and Krueger [8]. While the primary focus of their study did not revolve on the EKC, it was mentioned as a noteworthy discovery. Based on the EKC hypothesis, an inverse U-shaped association exists between some environmental indicators and economic advancement, as measured by per capita income. The proposition posits that an inverse relationship exists between a nation’s income and its environmental condition, wherein environmental degradation tends to worsen during the earliest periods of economic expansion, but above a certain threshold of economic prosperity, environmental quality begins to ameliorate. It is imperative to acknowledge that subsequent to the initial discovery by Grossman and Krueger [8], the EKC has garnered significant attention and scrutiny within the realm of environmental economics. The results obtained from conducting experiments on the notion, while manipulating various environmental factors and income levels, have shown inconsistent findings. There are several factors that might potentially influence the relationship between economic growth and environmental quality, including the type of pollutant, advancements in technology, government interventions, and regional variations. Consequently, EKC does not possess the universal applicability of a theory, but rather serves as a concept that has stimulated more investigation and discourse within the field. Nevertheless, the empirical data regarding the EKC are varied and contingent on the specific setting [9].

There is a close connection among pollution, energy use, and economic expansion [10]. Given the escalating pace of global warming, it is necessary to investigate the fluctuations in pollutant levels in correlation with variations in gross domestic product (GDP). The authors investigates the existence of an EKC across many panels of nations categorized into low-, middle-, and high-income status groups globally from 1990 to 2016. To evaluate EKC in connection to carbon dioxide, nitrous oxide, and methane, a fixed/random effect regression analysis is performed using robust standard error estimation, as determined by the Hausmann test. The Kaya identity is employed in a suitable manner to analyze the increase in contributions of different component ratios to global pollution rates across many countries. The panel vector error correction model was used to analyze the unidirectional and bidirectional causal connections between the gross domestic product (GDP) and pollutant emissions.

Dogan and Ozturk [11] examined the effects of real income and the use of both renewable and non-renewable energy on CO₂ emissions in the United States from 1980 to 2014. They employed the EKC model as their analytical framework. The findings from the Gregory–Hansen cointegration test suggest that there is an identified long-term connection among CO₂ emissions, real income, quadratic real income, and both renewable and non-renewable energy in use. The long-term estimates of the autoregressive distributed lag (ARDL) model suggest that a rise in renewable energy (RE) consumption has a mitigating effect on environmental damage, whereas an increase in non-renewable energy consumption contributes to carbon dioxide (CO₂) emissions. Furthermore, the EKC theory lacks validity when applied to the United States.

The examination and discussion of the relationship between economic expansion and environmental deterioration, particularly in terms of pollution, have been prominent areas of scholarly investigation and policy discourse. Historically, there has been a correlation between economic development, as measured by a rise in GDP and greater earnings, and the enhancement of living standards and overall prosperity. Nevertheless, it is crucial to acknowledge the significant correlation between economic growth and heightened environmental degradation, namely in terms of pollution [9]. The complex correlation between economic success and environmental pollution spans several dimensions and is shaped by a range of factors, such as advancements in technology, interventions in policy, and the level of social consciousness [12].

Industrialization and urbanization are key mechanisms via which economic growth exerts its influence on pollution. The process of economic development frequently entails a transition from agrarian-oriented economies to economies that are centered around industrial and service sectors. This period of transition is marked by a notable surge in industry, energy use, and transportation endeavors. The activities release several pollutants, including sulfur dioxide (SO₂), nitrogen oxides (NOx), and particulate matter, hence contributing to the occurrence of air pollution [13]. The EKC posits a curvilinear association between income and pollution, characterized by an initial rise in pollution levels during the early stages of economic expansion, followed by a subsequent decline as societies achieve higher levels of wealth [8].

Economic growth frequently triggers the process of urbanization and results in heightened population density. However, if this urbanization is not accompanied by efficient urban planning, it can give rise to elevated levels of pollution. The process of urbanization occurring at a fast pace has the potential to lead to heightened levels of energy consumption, traffic congestion, and industrial activity, all of which have been identified as factors that contribute to the problem of pollution [14]. Nevertheless, urban planning assumes a pivotal role in the mitigation of these adverse environmental effects. The implementation of well-structured urban planning strategies has the potential to facilitate the development of sustainable transportation systems, green infrastructure, and effective land utilization. Consequently, these measures can contribute to the mitigation of pollution levels in rapidly expanding metropolitan areas [15].

Increases in economic activity stimulate transportation activity, resulting in elevated levels of pollution. The escalation of earnings leads to a heightened desire for autos and transportation services, thus leading to an upsurge in the discharge of greenhouse gases (GHGs) and air pollutants [16]. Strategies aimed at alleviating this impact encompass the allocation of resources towards the development and implementation of cleaner transportation technology, as well as the promotion of sustainable urban design practices with the objective of diminishing dependence on private vehicles [17].

As civilizations experience economic growth and an increase in affluence, there tends to be a corresponding inclination towards higher levels of resource consumption, resulting in a subsequent rise in waste generation [18]. To address the escalating environmental consequences resulting from heightened waste generation in expanding economies, it is imperative to implement sustainable waste management approaches, including recycling, circular economy concepts, and trash reduction initiatives [19].

The extraction and utilization of the natural resources, such as minerals, fossil fuels, and agricultural products, frequently lead to environmental damage and pollution. As exemplified by Dasgupta [18], mining activities have the potential to result in adverse environmental consequences, such as soil and water contamination, deforestation, and habitat damage. Industries that require substantial resources, such as energy generation and manufacturing, make a substantial contribution to environmental pollution. The aforementioned industries not only exhibit high levels of resource consumption, but also contribute to the release of greenhouse gases (GHGs) and other harmful pollutants into the environment [20].

As countries experience economic growth, there is a tendency for them to allocate greater resources towards research and development (R&D), resulting in advancements in environmentally sustainable energy sources, enhanced transportation infrastructure, and the adoption of eco-conscious manufacturing techniques [21]. Furthermore, the presence of economic growth can lead to the enforcement of more stringent environmental rules. These regulations serve as incentives for firms to decrease their emissions and adopt cleaner manufacturing techniques [22].

Economic expansion frequently serves as a catalyst for technological advancements, which in turn can exert an influence on levels of pollution. This, in turn, facilitates advancements in the realm of sustainable energy sources, environmentally conscious manufacturing techniques, and enhanced pollution mitigation technology [21]. Technological improvements have the potential to result in a decrease in pollution emissions per unit of economic output, a phenomenon sometimes referred to as “dematerialization”. According to Jorgenson and Clark [23], scholarly research has demonstrated that the advancement of technology has a substantial influence in disassociating economic expansion from pollution.

The relationship between higher earnings and economic development has the potential to foster heightened environmental consciousness and a notable movement in consumer preferences towards sustainable practices. As individuals and communities experience an increase in income, there is a tendency to place greater importance on environmental conservation and exhibit a higher demand for items and behaviors that are environmentally friendly [9].

The presence of pollution can result in negative consequences for human well-being, ultimately resulting in heightened healthcare costs [24]. In addition, pollution has the potential to disturb ecosystems, diminish agricultural output, and contribute to climate change, as highlighted by Costanza et al. [20]. The presence of negative externalities has the potential to impede the sustained expansion and advancement of the economy, hence emphasizing the significance of tackling pollution as a crucial economic and environmental obstacle.

The complex interplay between economic growth and pollution presents considerable obstacles for policymakers. Achieving a harmonious equilibrium between promoting economic growth and minimizing negative environmental impacts necessitates the implementation of a comprehensive and complex strategy. This entails formulating rules that facilitate the adoption of sustainable practices, provide incentives for the utilization of cleaner technologies, and establish punishments for instances of environmental harm. In addition, it is imperative to emphasize the need for international collaboration in addressing global environmental concerns, as pollution frequently surpasses national boundaries, necessitating collective endeavors [25].

The impact of economic expansion on pollution is further moderated by governmental environmental restrictions. The primary objective of these policies is to alleviate the environmental consequences associated with economic operations. Research has indicated that the implementation of rigorous environmental regulations can successfully mitigate pollution levels, even in the presence of economic growth [26]. According to the environmental Kuznets curve theory, a positive relationship exists between economic growth and pollution levels in the early stages. However, once a particular income threshold is surpassed, the implementation of environmental legislation and increased awareness likely result in a decline in pollution levels [9].

Hamaguchi et al. [27] examine how environmental policies influence economic growth and pollution. The theory that implementing stringent environmental regulations will lead to a decline in employment opportunities and reduced levels of productivity is deeply entrenched. To promote sustainable development, it is crucial to ascertain environmental policies that align with both economic growth and the mitigation of pollution.

Onofrei et al. [28] investigates the relationship of economic development and carbon dioxide (CO₂) emissions across the 27 member states of the European Union (EU) over the period spanning from 2000 to 2017. The authors utilize a combination of quantitative methodologies, including dynamic ordinary least squares (DOLS), unit root tests, and cointegration techniques, alongside a qualitative sequential methodology that incorporates empirical analysis. The findings suggest that a durable relationship exists between economic development and CO₂ emissions in European Union (EU) nations. The dynamic ordinary least squares approach demonstrates that economic growth has a statistically substantial influence on CO₂ emissions, as indicated by both versions of estimators. Specifically, the analysis reveals that, on average, a 1% alteration in GDP leads to a 0.072 unit change in CO₂ emissions.

In a study conducted by Zoundi [29], an examination is made into the effects of renewable energy on carbon dioxide (CO₂) emissions testing on a panel of 25 countries from Africa. The investigation spans the years from 1980 to 2012 and employs the panel cointegration technique. It has been observed that there is a positive correlation between economic expansion and the rise in CO₂ emissions. The estimation findings show that renewable energy has a negative impact on CO₂ emissions, making it an effective alternative to existing energy sources. African countries are amongst the heaviest releasers of pollution due to old technology and a lack of investments.

Ozturk and Acaravci [30] investigated the connection among macro-economic variables, economic growth, foreign trade, employment ratio, and energy specific indicators, such as carbon emissions and energy consumption, in Cyprus and Malta from 1980 to 2006. They employed the autoregressive distributed lag (ARDL) bounds testing approach and error correction-based on Granger causality models to analyze the data. The empirical evidence suggests that there is a long-term association between the variables; however, this relationship is only observed in the case of Malta. Still, the countries are not representative at the European level when discussing heavy industry and the level of pollution.

Bovenberg and Smulders [31] examined the correlation between economic expansion and environmental quality by including a pollution-increasing technology breakthrough in an endogenous development model. This study examines the factors that contribute to the feasibility and desirability of achieving sustainable economic growth. The authors developed a comprehensive framework aimed at implementing a more assertive environmental policy with the objective of enhancing long-term economic growth.

There is a growing impetus in the endeavor to separate economic progress from pollution. The notion of sustainable development underscores the imperative of attaining economic growth while concurrently safeguarding the environment and maintaining social fairness [32]. The prioritization of environmental sustainability with economic success is a key feature of the sustainable development goals, as articulated in the United Nations’ 2030 Agenda for Sustainable Development [33].

In a recent study, Mitic et al. [34] studied the connection among CO₂ emissions, energy production and socio-economic variables, like increases in GDP and employment, in a panel of eight countries from southern and eastern Europe between 1995 and 2019. The research indicated that bidirectional causality exists between CO₂ emissions and employment as well as between energy and employment. Unidirectional causality was observed between employment and GDP.

Chomac-Pierzecka et al. [35] centers around the ongoing energy transition in European Union nations, specifically highlighting the formal regulations and the influence of decision-makers supported by the public. The authors explore the worldwide implementation of environmental conservation strategies and the broad adoption of the Green New Deal within the energy industry. The study investigates the level of familiarity across communities with mainstream transition solutions, such as renewable energy sources, and examines how public awareness impacts consumer choices, thereby altering the market. Predominantly carried out in Poland and Lithuania, the study examines these nations based on their close geographical proximity and comparable socio-economic circumstances, providing insights into the energy issue. The results indicate that heightened public consciousness promotes the acceptance of renewable energy alternatives, regardless of their specific nature, hence facilitating the process of transitioning to sustainable energy sources.

Jozwik et al. [36] explores the dynamic transformations brought about by environmentally friendly activities and the shift towards digital technologies in countries located in Central Europe. These projects incorporate cutting-edge technologies into everyday life and commercial operations, resulting in improved environmental quality through sustainable practices. The study seeks to elucidate the complex interrelationships among the utilization of renewable energy, digitalization, financial development, and their collective influence on environmental quality in the region. By analyzing data spanning from 1995 to 2019, the study reveals a positive correlation between economic growth and carbon emissions, while also identifying a negative correlation among digitization, adoption of renewable energy, and carbon emissions. The process of digitalization acts as an intermediary in the connection between the utilization of renewable energy and the state of the environment in nations, such as the Czech Republic, Hungary, Latvia, and Slovakia. Nevertheless, the impact of financial development differs between nations. These findings provide useful insights that can be used to shape policy suggestions in Central European nations.

The study prepared by Neagu [37] used the EKC model to examine the relationship between the economic complexity index (ECI) and carbon emissions in 25 European Union (EU) nations from 1995 to 2017. The study utilizes a cointegrating polynomial regression (CPR) to analyze both panel data and individual nation time series. The model also takes into account the concept of “energy intensity” as a crucial determinant of carbon emissions. According to the study, countries have an inverted U-shaped pattern in CO₂ emissions based on their economic complexity. Initially, pollution levels increase as countries expand their range of exported goods. However, after reaching a certain threshold, higher economic complexity results in decreased emissions. The analysis reveals a significant correlation among economic complexity, energy intensity, and carbon emissions. Specifically, a 10% increase in energy intensity leads to a 3.9% increase in CO₂ emissions over the long term. This quadratic model, which incorporates the economic complexity index (ECI), has been verified for all nations in the panel, including Belgium, France, Italy, Finland, Sweden, and the United Kingdom. The paper also examines the graphical depictions of the environmental Kuznets curve (EKC) in various nations and incorporates policy ramifications.

It is essential to reassess the EKC hypothesis in order to comprehend global economic progress and accomplish carbon reduction objectives, as an increasing number of nations pledge to achieve carbon neutrality and decrease emissions. The primary objective of this study [38] is to address the difficulty of pinpointing the exact moment when an increase in income results in a decrease in CO₂ emissions. The EKC hypothesis is investigated for G7 countries from 1890 to 2015 using a novel kink regression model. The findings suggest that the inverted U-shaped environmental Kuznets curve phenomenon is not applicable to the United States, Germany, Italy, Canada, and Japan. Instead, a “pseudo-EKC” pattern emerges, wherein the positive relationship between income and CO₂ emissions weakens at a certain threshold. The classic environmental Kuznets curve is only adhered to by the United Kingdom and France. Moreover, there is an inverse U-shaped correlation between wealth and SO₂ emissions, with several inflection points happening at varying times across different G7 nations. These findings indicate that while developing environmental strategies to reduce pollutants, it is important to take into account the distinct attributes of different contaminants and areas.

Through the review of a significant portion of studies from the specific literature analyzing the connection between CO₂ emissions and socio-economic variables, it was observed that there is a gap related to a higher period of observation. Thus, the current study proposes a larger timeframe for analysis, from 1970 to 2021, in order to capture the period of heavy industrialization of the former communist countries but also the dot-com bubble period, which was characterized by a significant impulse in the production of electronics and electric components. Both periods mentioned assumed a higher consumption of energy and increase in population based on the specific Soviet policies.

Three research hypothesis were formulated based on the literature review and the objective of the paper to illustrate the connection between the variables, specifically the influence of the independent variables on the dependent variable, CO₂ emissions.

H1. 

GDP/capita growth has a statistically significant impact, positive or negative, on CO₂ emissions.

H2. 

Population increase has a statistically significant positive influence on CO₂ emissions.

H3. 

Primary energy consumption has a statistically significant positive influence on CO₂ emissions.

3. Research Methodology and Data

The research aim is to determine if a direct relation exists among the following 3 variables in the evolution of CO₂ emissions: the gross domestic product per capita (GDP/capita), a macro-economic variable; the evolution of the population, a demographic variable; and energy consumption, an industrial variable.

The first hypothesis defined seeks to prove that the growth of GDP per capita has a statistically significant impact, either positive or negative, on carbon dioxide (CO₂) levels. This hypothesis seeks to examine the correlation between the GDP/capita and carbon dioxide (CO₂) emissions. The purpose of this study is to ascertain whether variations in a nation’s economic well-being, as measured by GDP per capita, have a statistically significant influence on CO₂ emissions. The hypothesis does not make any assumptions regarding the exact direction (positive or negative) of this impact as the objective is to determine if there is a correlation between economic growth and changes in CO₂ emissions, and if such correlation exists, to assess its significance.

The second hypothesis seeks to observe if population growth has a statistically significant and beneficial impact on the increase in CO₂ levels. This hypothesis investigates the impact of population expansion on CO₂ emissions and suggests that when the population of a country grows, there will be a statistically significant and beneficial impact on the emission of CO₂. Put simply, it implies that a greater population results in elevated levels of CO₂ emissions, most likely due to heightened energy usage, transportation demands, and industrial operations linked to a larger labor force and increasing number of consumers.

The third hypothesis aims to prove if primary energy use has a considerable positive impact on CO₂ emissions. This hypothesis examines the correlation between the use of primary energy and the release of CO₂ emissions. This hypothesis suggests that when a country’s main energy consumption rises, there will be a statistically significant correlation with an increase in CO₂ emissions. Put simply, this hypothesis proposes that increased energy consumption, namely from non-renewable sources, results in higher levels of CO₂ emissions. This is because the combustion of fossil fuels is a significant contributor to carbon emissions.

To summarize, these assumptions establish a structure for an empirical investigation into the effects of GDP per capita growth, population growth, and primary energy use on CO₂ emissions. The objective is to ascertain the presence of substantial correlations between these parameters and CO₂ emissions and to determine the direction of the link for H1 and H3.

The study involves the collection of data on four variables across the time span from 1970 to 2021, with 50 yearly observations available for each indicator. The study employed empirical analysis to ensure and establish coherence, including techniques such as unit root testing and regression analysis. After conducting a comprehensive literature search, it was determined that panel data analysis is appropriate for this study due to the inclusion of numerous countries with data for each measurement in the analysis.

The variable of economic growth is closely monitored and studied and is often measured by gross domestic product per capita. This indicator is considered representative as it reflects the economy’s ability to produce goods and services, which in turn leads to a significant release of CO₂ into the atmosphere. The population level was taken into account due to the inherent human need for goods and services, resulting in the release of a certain amount of CO₂ into the atmosphere. Additionally, energy consumption is closely linked to both economic progress and population growth, as the latter requires energy in various forms (

Table 1. The variables used.

Variable	Symbol of the Variable	Unit of Measurement	Source of Data
Gross domestic product per capita level	GDP/capita	Thousands USD, 2011 prices, real terms	Maddison Project Database 2020 [39], Eurostat [40]
Population	POP	Millions	Our World in Data [41]
Energy consumption	ENERG	Terawatt-hours per year	U.S. Energy Information Administration (EIA) [42], Energy Institute Statistical Review of World Energy (2023) [43]
CO2 emissions	CO2	Million tonnes	Global Carbon Budget (2022) [44]

).

The study proposes a quantitative approach to determine the connection between the variables representing the real economy, demography of the countries and the energy used, and the quantity of CO₂ emissions. Descriptive statistics were conducted to proper represent and analyze the dataset (

Table 2. Descriptive statistics.

Indicator/Variable	CO2	Population	GDP/Capita	Energy Consumption
Mean	171.6185	20.11975	24.75066	818.2163
Median	78.93000	10.20000	23.00305	410.7250
Maximum	1117.890	83.41000	113.1824	4382.910
Minimum	1.700000	0.340000	3.311953	8.010000
Std. Dev.	217.2172	21.92407	13.19779	978.0074
Skewness	2.387285	1.452215	1.225327	2.004589
Kurtosis	8.918177	3.938479	6.875761	6.438859
Jarque-Bera	2505.592	403.7130	911.1799	1208.968
Probability	0.000000	0.000000	0.000000	0.000000
Observations	1040	1040	1040	1040
Cross-sections	20	20	20	20

).

For testing the hypothesis, a panel data model was built by using a group of 20 countries from the European Union: Austria, Belgium, Bulgaria, Republic of Cyprus, Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Luxembourg, Netherlands, Poland, Portugal, Romania, Spain, and Sweden. The testing was realized on 52 annual observations from 1970 to 2021, a time interval considered sufficiently large to include at least three moments of economic downturns. Countries like Croatia, Estonia, Latvia, Lithuania, Malta, Slovakia, and Slovenia were not included in the study due to lack of data.

The data were collected from public sources and included values for all the variables defined in the research for all 20 countries on which the panel data were built. Input data are represented by the four variables defined in the previous paragraphs for each of the 20 countries considered for analysis.

Baltagi [45], Hsiao [46], and Klevmarken [47] provide a comprehensive analysis of the advantages and drawbacks associated with the utilization of panel data models. The benefits that have been found allows one to observe the following arguments: (1) The utilization of panel data enables an enhanced and efficient utilization of information by combining cross-sectional and time series data, thereby increasing statistical power. (2) Panel data analysis offers the ability to control for unobserved heterogeneity across individual units through the application of fixed and random effects models, thereby enhancing the precision of estimations. (3) The utilization of panel data facilitates the examination of dynamics and temporal changes by capturing individual trajectories and trends. (4) Panel data analysis enables improved identification of causal relationships and policy effects by effectively controlling for unobserved variables and individual-specific effects. Panel data approaches frequently result in more accurate estimates in comparison to using only cross-sectional or time series data, hence improving efficiency and decreasing the variance of the estimator.

The panel data models have been recognized to have some limitations. First, the collection of consistent and comprehensive panel data over time might pose challenges due to the associated expenses of data collection. Additionally, panel attrition and the possible loss of observations or participants further contribute to the difficulties faced in obtaining reliable panel data. The phenomena of panel attrition and selective participation have the potential to introduce selection biases, which can undermine the representativeness of the panel and potentially distort the estimation results. Panel data frequently exhibit unobserved heterogeneity, such as unmeasured individual characteristics, which can complicate the estimation of models and the interpretation of results. The analysis of panel data necessitates the use of sophisticated models and specifications, which increases the risk of model misspecification and the potential challenges in selecting an appropriate model.

3.1. Panel Unit Root Test

Panel unit root tests play a crucial role in the examination of the time series characteristics of variables within a panel dataset, considering both the individual and temporal aspects. The purpose of these tests is to evaluate if the variables being analyzed have a unit root, which suggests non-stationarity. The Levin-Lin-Chu (LLC) test [48], which is an extension of the Dickey and Fuller [49,50] test for panel data, is a commonly employed panel unit root test. The LLC test incorporates individual-specific intercepts and trends, therefore accommodating cross-sectional variability also through Im-Pesaran and Shin test-IPS [51] and Fisher tests. The augmented Dickey-Fuller (ADF) unit root test was developed by Maddala and Wu [52] and Choi [53] and is demonstrated using the following equation:

Δy_it = α_i + β_it + γ_yi,t−1 + δ₁Δy_i,t−1 + δ₂Δy_i,t−2 + …+ δ_p−1Δy_{i,t−(p−1)} + ϵ_it

(1)

were Δy_it is the first difference of the variable for individual i at time t; α_i represents the individual-specific intercept; β_i is the individual-specific trend; γ measures the coefficient of the lagged level of the variable; δ₁, δ₂, …, δ_p−1 are coefficients for lagged differences; and ϵ_it is the error term.

The null hypothesis of the ADF test is that γ = 0, suggesting the presence of a unit root (non-stationary series). The alternative hypothesis is that γ < 0, indicating stationarity.

In order to perform the LLC test, a panel regression model is constructed for each variable, encompassing a temporal trend, lagged levels, and lagged differences. The estimation of critical parameters is conducted using regression analysis, and the LLC test statistic, which is a standardized t-statistic, is calculated based on these estimated parameters. The test statistic is subsequently compared to crucial values derived from the LLC distribution to ascertain if the null hypothesis, which posits the presence of a unit root and hence non-stationarity, may be rejected in favor of the alternative hypothesis of stationarity.

To apply the unit root test, first the hypotheses of the study was formulated. The null hypothesis (H0) indicates the existence of a unit root in the series, meaning that there is a non-stationery series of data, while the alternative hypothesis (H1) states that the series is stationery. The results are presented in

Table 3. Unit root tests.

Tests/Variables	CO2		Population		GDP/Capita		Energy Consumption
	Level	1st Difference	Level	1st Difference	Level	1st Difference	Level	1st Difference
LLC	0.2068	0.0000	0.0549	0.0005	0.4237	0.0000	0.0000	0.0000
IPS	0.8558	0.0000	1.0000	0.0000	0.9973	0.0000	0.0112	0.0000
ADF-Fisher	0.8727	0.0000	0.0874	0.0000	0.4332	0.0000	0.0706	0.0000
PP-Fisher	0.9042	0.0000	0.0659	0.0000	0.9986	0.0000	0.0635	0.0000

.

To strengthen the empirical analysis, the panel cointegration test was performed to observe the connection between the variables. The Pedroni [54,55] panel cointegration test has significant importance as a foundational analytical method for examining cointegration patterns within a panel dataset, which is extensively employed in the field of econometrics. The proposed methodology expands upon conventional cointegration analysis to incorporate cross-sectional dependency and heterogeneity, which are essential considerations in panel datasets. The assessment of long-term correlations among variables is conducted by the examination of the stationarity of the calculated coefficients in the test. The software provides a range of test statistics that are specifically designed to analyze different types of cointegration patterns [51,54,55]. The results are presented in

Table 4. Panel cointegration test.

Dimension	Statistic Test	Statistic	Probability
Within-dimension	Panel v-Statistic	0.928168	0.1767
	Panel rho-Statistic	−1.30877	0.0953
	Panel PP-Statistic	−2.087679	0.0184
	Panel ADF-Statistic	−1.889709	0.0294
Between-dimension	Group rho-Statistic	0.774924	0.7808
	Group PP-Statistic	−0.073759	0.4706
	Group ADF-Statistic	0.378061	0.6473

Note: Null hypothesis: No cointegration. Trend assumption: No deterministic trend. Individual intercept.

.

Further, the estimation of the panel regression was made for each variable and based on Equation (1). This step is followed by the calculation of the LLC statistic. The LLC test statistic is a standardized t-statistic for testing the null hypothesis (H0) against the alternative hypothesis (H1). The values obtained are compared with the critical values from the LLC distribution to determine statistical significance.

3.2. Estimation Model—Panel Data Regression

Panel data refer to a statistical model that encompasses variables that exhibit variation over both time and cross-sectional units. The analysis employed the EGLS approach to estimate the unknown parameters of the linear regression model, taking into account a specific level of correlation among the residuals in the regression model. The generalized least squares (GLS) approach was initially introduced by Aitken [56]. It involves the estimation of a single equation that incorporates all variables, including cross-sectional and time series data, which are combined into a single column. The assessment of this equation is often conducted using the ordinary least squares (OLS) technique.

Generalized least squares (GLS) was used as a potential remedy for the issue of autocorrelation or heteroscedasticity, which violates the underlying assumption of ordinary least squares (OLS) that the error terms are uncorrelated and possess constant variance. Consequently, this violation renders the Gauss–Markov theorem inapplicable, thereby undermining the status of OLS estimators as best linear unbiased estimators (BLUE) and causing the resulting estimates to be inefficient. The rationale for employing generalized least squares (GLS) arises from the existence of autocorrelation or heteroscedasticity in the error terms.

Greene [57] and Wooldridge [58] state that pooled generalized least squares (pooled GLS) is usually employed to estimate the parameters in the linear regression model, while accounting for the analysis of pooled cross-sectional and time series data. In the context of pooled data, there exists a combination of cross-sectional units, such as persons or nations, together with many time periods. The pooled generalized least squares (GLS) method enables the estimation of parameters by taking into account any correlations and heteroscedasticity present within the dataset. The objective of the pooled generalized least squares (GLS) estimator is to minimize the total sum of squared errors, while also accounting for the presence of heteroscedasticity and serial correlation. The assumed structure of the covariance matrix for the error element is commonly calculated using established techniques, such as heteroscedasticity-robust standard errors or autoregressive structures. The estimation procedure entails the multiplication of the ordinary least squares (OLS) estimator by the inverse of the square root of an estimated covariance matrix of the errors. This process aids in the transformation of mistakes, resulting in homoscedasticity and uncorrelatedness, which in turn leads to improved efficiency in estimate. The pooled GLS estimator (β^PGLSβ^PGLS) is computed as follows:

β_PGLS = (X^′Ω⁻¹X)⁻¹X^′Ω⁻¹y

(2)

where X is the design matrix of independent variables, y is the vector of the dependent variable, and Ω is the estimated covariance matrix of the error term.

The purpose of the regression equation is to assess the association between a dependent variable and independent variables, while considering both cross-sectional and time series aspects. This methodology integrates the advantages of data pooling while simultaneously addressing the issues of heteroscedasticity and potential serial correlation. The general form of the equation can be written as follows:

Y_it = β₀ + β₁X_1it + β₂X_2it +…+ β_kX_it + u_it

(3)

where Y_it is the dependent variable for the ith unit at time t; X_1it, X_2it, …, X_kit are independent variables for the ith unit at time t; β₀, β₁,…, β_k are the coefficients to be estimated; and u_it is the error term for the ith unit at time t.

The application of Equation (3) to the current research will yield the following equation:

CO_2it = β₀ + β₁GDP/capita_it + β₂POP_it + β₃ENERG_it + u_it

(4)

The equation results are presented in

Table 5. Equation estimation.

Variable	Coefficient	t-Statistic	Prob.
GDP/capita	−1.621118	−140.3957	0.0000
Population	0.743761	28.07425	0.0000
Energy consumption	0.199582	235.6932	0.0000
R-squared			0.997013
Adjusted R-squared			0.997004
F-statistic			115,252.8
Prob(F-statistic)			0.0000

.

Seemingly unrelated regression (SUR) [59] proposes an approach to estimate panel data models that include a big number of time periods (T) but a small number of cross-sectional units (N) and is known as the “long T, small N” method. To conduct estimation of a seemingly unrelated regression (SUR) model, it is necessary to organize the data in the form of a time series rather than a panel dataset, with distinct variables listed individually. The fundamental seemingly unrelated regression (SUR) model assumes that the errors exhibit homoscedasticity and linear independence inside each equation. Each equation may have its own variance. Each equation exhibits a correlation with the other equations within the same temporal interval. The last assumption is referred to as contemporaneous correlation, and it is this characteristic that distinguishes SUR from other models.

3.3. Panel Vector Autoregression (VAR)

Panel vector autoregression (VAR) models [60,61] are a reliable statistical framework used in empirical research, specifically for examining the dynamic connections between several time series variables in a panel dataset. The models enhance classic VAR analysis using panel data, which involves collecting observations over both time periods and cross-sectional units. The use of panel VAR models excels in capturing both the individual-specific and common dynamics, enabling the analysis of the evolution of variables inside each entity and the examination of potential variations in these dynamics across entities.

4. Empirical Results

The analysis started by applying the descriptive statistics for the variables used in order to have a broader picture of the data used in analysis. Each variable has been observed 1040 times across 20 different cross sections. The standard deviation for carbon dioxide emissions is approximately 217.22, indicating there is a high degree of variability in the dataset which indicates the different degree of industrialization but also different population levels. There is a rightside skewed distribution for all the variables analyzed while the high positive kurtosis indicates a distribution with heavier tails and a slightly sharper peak. The Jarque-Bera test for normality evidenced that the null hypohotesis of normality was rejected.

The methodology presented starts with the unit root test for the panel of data to check for the stationarity of the series and unit roots. The tests performed were proposed by Levin et al. [48] and Im et al. [62] but also the newer version of the tests, such as ADF-Fisher and PP-Fisher proposed by Maddala and Wu [52] and Choi [53]. The latter proposed a Fisher-type test for examining the presence of a unit root in panel data. This test involves combining the P-values obtained from unit root tests conducted on each cross-section i. The Fisher test is a nonparametric statistical test that follows a Chi-square distribution with two degrees of freedom.

The presence of a unit root was examined through hypothesis testing, both at the series level and after taking the first difference. The results indicate that the series is not stationary, as determined by evaluating the probability associated with the t-statistic for the first difference. This analysis was conducted with confidence levels of 1%, 5%, and 10%. The results of the panel unit root test suggest that there are unit roots in the levels of CO₂, population, and GDP per capita. However, after taking the first difference, all variables exhibit stationarity, indicating that the data series is integrated at order one. Furthermore, the results of the cointegration test indicate that the null hypothesis assuming no cointegration is rejected when including the intercept for three tests if we consider a 10% significance level and two tests at a 5% significance level. We observe that there is a good level of testing, and the cointegration of the variables was proved by the analysis.

The unit root test results show that the data do not possess a unit root and that the time series is stationary. Stationary time series display a consistent mean and variation throughout the whole time period. The time series chosen is amenable to modeling and analysis because of the absence of long-term trends or seasonal patterns.

The results of the panel cointegration test show that cointegration exists and that a durable connection exists between the variables in the panel dataset. In this scenario, the variables exhibit a long-term movement in unison, indicating the presence of a shared underlying trend or equilibrium.

4.1. Regression Results

Based on the results obtained during the unit root test indicating that the series is not stationary, an empirical analysis was continued with the pool estimation of the regression equation. The test used was pool estimation based on least squares method and cross-section SUR. The dependent variable was CO₂, while the independent variable were population, GDP/capita, and energy consumption.

The least squares approach is a widely employed statistical analysis methodology that aims to estimate the parameters of a linear model [63,64]. This is achieved by minimizing the sum of squared differences between the observed values and the values predicted by the model. The objective is to identify the line that provides the most accurate representation of the association between variables. From a mathematical perspective, the objective is to minimize the sum of squared residuals or mistakes. The estimated coefficients obtained from the analysis are considered ideal since they minimize the error to the greatest extent feasible.

From a technical perspective, the equation built to estimate the coefficients is valid considering that the probability associated with the t-statistic is accepted at confidence levels of 1%, 5%, and 10%. Furthermore, the probability associated with the f-statistic is relevant at a confidence level of 1%, 5%, and 10%, indicating that the regression equation is relevant.

4.2. VAR Results

For a better observation of the quantitative analysis, a vectorial VAR was applied in order to test the correlation between variables and to reflect their effects. The impulse response functions were determined using all four variables defined in the study (

Table 6. Results of the impulse response functions of 4 variables, fist iteration.

Response of		Response to
	CO2 (t − 1)	Population (t − 1)	GDP/Capita (t − 1)	Energy Consumption (t − 1)
CO2 (t)	10.92540 (0.24430)	0.000000 (0.00000)	0.000000 (0.00000)	0.000000 (0.00000)
Population (t)	0.005117 (0.00094)	0.029448 (0.00066)	−0.000379 (0.00093)	0.000000 (0.00000)
GDP/capita (t)	315.7200 (58.5520)	0.000000 (0.00000)	1838.068 (41.1005)	0.000000 (0.00000)
Energy consumption (t)	34.08416 (1.09227)	−0.829107 (0.78097)	1.388394 (0.78181)	24.68955 (0.55208)

and

Table 7. Results of the impulse response functions of 4 variables, second iteration.

Response of		Response to
	CO2 (t − 2)	Population (t − 2)	GDP/Capita (t − 2)	Energy Consumption (t − 2)
CO2(t)	10.64347 (0.42431)	0.068310 (0.09772)	0.772373 (0.35828)	0.185356 (0.34668)
Population (t)	0.017732 (0.00210)	0.057551 (0.00131)	0.001486 (0.00207)	0.000652 (0.00095)
GDP/capita (t)	409.8993 (87.0240)	5.156797 (16.7757)	1976.791 (75.4236)	−64.92757 (59.1932)
Energy consumption (t)	35.51604 (1.74039)	−0.087613 (0.83613)	4.008541 (1.57113)	23.61657 (1.43694)

).

We observe that the response of CO₂ emissions to population is positive in the estimated coefficients and impulse response, in accordance with the economic theory that a larger population will produce more CO₂ emissions. The response of the coefficients of CO₂ emissions to GDP/capita and energy consumption are even higher with an important degree of significance showing that an increase in economic activity will lead to an increase in gas emissions.

The variance decomposition for the VAR model presented in

Table 8. Variance decomposition.

Variable	CO2	Population	GDP/Capita	Energy Consumption
10 periods ahead
CO2	100.0000	0.146138	0.308764	0.108767
Population	15.48094	97.05294	0.432352	0.219576
GDP/capita	4.122420	0.003775	97.13415	0.346205
Energy consumption	70.34004	0.425826	0.642945	34.36312
20 periods ahead
CO2	100.0000	0.586779	0.546847	0.639156
Population	97.05294	97.05294	0.506121	0.728044
GDP/capita	4.158571	0.004578	97.13415	0.897355
Energy consumption	71.55037	1.770678	0.642945	34.36312

is in line with the previous findings of the paper. The GDP/capita, which is actually an indicator of an advanced economy, contributes up to 30% to CO₂ emissions, while population and energy consumption contribute between 10% and 15%. After 20 periods, the contribution of the energy consumption becomes dominant and can explain up to 64% of CO₂ emissions, whereas population and GDP/capita can explain up to 59%.

5. Discussion

The research objective was to determine the relation among specific economic, demographic and energy variables on the environmental variable using empirical analysis through collection of data and specific statistical tests. The macro-economic variable used was the GDP per capita as it shows the level of economic development not just as a wide measure but also by considering the country’s population. Specifically, the greater the GDP per capita, the greater the economic development. On the demographic side, the population level was chosen considering that the higher the population in a country, the higher the needs for energy and goods and services for which the energy is used. The third variable chosen was the level of energy consumption as this shows two relevant aspects, namely, the use of energy by the population but also by industry, which is the biggest consumer. All the variables mentioned were measured to assess their influence on CO₂ emissions.

The results of the unit root test proved that the series analyzed is not stationary and the null hypothesis was rejected at the first difference. Specifically, the panel was integrated at order one.

Arouri et al. [65] found similar results in their study demonstrating that over an extended period, energy consumption exerts a favorable and substantial impact on CO₂ emissions. Significantly, it illustrates that there is an intricate, quadratic correlation between real GDP and CO₂ emissions throughout the region. Although the majority of countries exhibit a correlation between long-term income coefficients and their squares that supports the notion of the environmental Kuznets curve (EKC), the specific turning points at which this correlation occurs vary significantly. This variation in turning points offers mixed evidence for the hypothesis of the EKC. Significantly, the MENA (Middle East and North Africa) area successfully decreased its per capita CO₂ emissions while also undergoing economic expansion between 1981 and 2005.

Bastola and Sapkota [66] measured the correlations among energy consumption, pollution emissions, and economic growth in Nepal by employing Johansen cointegration and ARDL bounds tests. Similar to the present research, the study revealed the presence of two cointegrating vectors, including one relating to energy consumption and the other pertaining to carbon emissions, as dependent variables. The Granger causality used in the study revealed a reciprocal causative relationship between energy consumption and carbon emissions, indicating that alterations in one variable affect the other. Additionally, there is a one-way causal relationship from economic growth to both carbon emissions and energy consumption. These findings suggest that strategies focused on boosting energy consumption may not necessarily foster economic growth and are more likely to have negative environmental consequences. In contrast, measures that encourage energy conservation and the reduction of carbon emissions do not impede long-term economic growth. Hence, directing attention towards alternative energy sources could prove to be a successful strategy in reducing the adverse effects of climate change and safeguarding the environment, all while ensuring the nation’s sustained economic development.

Similar results were observed by studying the economies of Turkey and Georgia [67]. The main finding of this analysis is that there is a significant correlation among carbon emissions, energy consumption, and economic growth in both the Turkish and Georgian economies. Crucially, the test results support the well-established environmental Kuznets curve hypothesis, indicating that as economies expand, they first lead to increased emissions but eventually show the capacity for emissions reduction through further progress.

Based on the R-squared statistical test, it was revealed that the model built through the regression equation is fit for the research as it indicates that the evolution of the independent variables explains up to 99.7% from the evolution of the dependent variable.

The results obtained through the empirical analysis validate the hypothesis, indicating that the coefficients are significant. There is a direct and positive correlation with the level of population as it explains up to 74.3% of CO₂ emissions. This finding is in line with the economy theory as the people are requiring goods and services that are produced with a CO₂ footprint but also considering the individual use of energy and their sources of production. Furthermore, the consumption of energy also exhibits a direct and positive influence on the evolution of CO₂ emission and explains up to 19.9% of the evolution based on the model. Considering that the vast majority of sources of energy production result in pollution, the hypothesis is accepted, and the direct influence of energy consumption on CO₂ emissions was demonstrated. Regarding macro-economic variables, a negative direct correlation exists as observed through the model. Economic theory could explain this finding as the higher the degree of the development of a country, the more attention is given to the green energy and environmental policies.

A study on South Asian countries [68] revealed a notable decrease in the quality of the environment, indicating a deterioration in environmental circumstances. Nevertheless, in this particular area, the environmental Kuznets curve (EKC) theory is substantiated by a blend of both declining and increasing levels of economic growth (GDP) and the square of GDP. This implies that as economies expand, they initially contribute to an increased in emissions, but subsequent phases of development can result in a decrease in emissions. Similar to the current research, the study expounds the relationship among globalization, energy use, and economic growth in South Asian countries. The evidence suggests that although there are obstacles regarding CO₂ emissions, the EKC hypothesis shows potential for this area, and a dynamic correlation exists between economic expansion and energy use.

In a similar study, Favero et al. [69] explored the worldwide correlation between economic growth and CO₂ emissions. The study employs a novel quantitative methodology, encompassing data from 187 nations during the time period from 1800 to 2016. The proposed multilevel model considers the complex relationships between fixed and random effects factors related to a country’s GDP and carbon dioxide emissions. The analysis not only validates a strong and statistically significant correlation between economic growth and carbon emissions within each country, but also uncovers novel findings. The analysis underscores the favorable influence of linear and cubic income on CO₂ emissions, while highlighting the detrimental impact of quadratic income. This underscores the validity of the typical N-shaped curve observed throughout the period under study. Crucially, this research demonstrates that these correlations remain constant in both rich and developing nations, while there may be differences across specific countries. These findings have important implications for researchers and policymakers who are studying the impact of economic growth on CO₂ emissions. They highlight the importance of implementing comprehensive policies to solve this global concern. Furthermore, the results were similar to the findings of Osobajo et al. [70], which showed that population, capital stock, and economic growth have a bidirectional connection with CO₂ emissions as observed in a panel of 70 countries between 1994 and 2013 by applying a pooled OLS regression.

The investigation disclosed a substantial and favorable correlation between energy usage and carbon emissions. Consequently, a rise in energy consumption is directly linked to an increase in carbon emissions. The statistical tests demonstrated that the regression model was highly suitable for the study, as it effectively accounted for a significant portion of the variability in carbon emissions, based on the independent variables. The study’s findings confirmed the theory, demonstrating a direct association between population and CO₂ emissions, a direct connection between energy consumption and CO₂ emissions, and an inverse correlation between GDP per capita and CO₂ emissions. These findings align with current research regarding the correlation between economic progress and the release of CO₂ emissions.

The contribution of the study to the literature includes demonstrating the correlation between CO₂ emissions and socio-economic variables over a longer timeframe and also testing the period of heavy industrialization in the former Soviet countries. Furthermore, the study provides a new view of the correlation between the GDP per capita and CO₂ emissions, and the negative connections demonstrate the increased importance of sustainability issues and related policies.

6. Conclusions

Through this study and empirical analysis, three fundamental hypotheses were investigated to gain a deeper understanding of the correlation between economic factors and carbon dioxide (CO₂) emissions. The first hypothesis, which sought to establish a connection between a country’s economic prosperity, as shown by GDP per capita, and CO₂ emissions, was validated as the study revealed a clear and inverse correlation. The results validated the correlation between fluctuations in GDP per capita and shifts in CO₂ emissions, highlighting the substantial influence of economic expansion on environmental consequences.

The second hypothesis, which examined the correlation between population growth and CO₂ emissions, posited that an increase in a country’s population would result in a statistically significant and direct impact on CO₂ emissions. The analysis unveiled a robust association, substantiating the idea. An expanded population results in elevated CO₂ emissions as a consequence of heightened energy requirements, transportation demands, and industrial operations, illustrating the impact of a burgeoning labor force and consumer population on environmental variables.

The third hypothesis examined the correlation between primary energy usage and CO₂ emissions. The hypothesis suggests that an increase in a country’s primary energy consumption will have a statistically significant beneficial effect on CO₂ emissions. The investigation validated this theory, underscoring the robust correlation between escalated energy use, specifically from non-renewable sources, and elevated CO₂ emissions. The burning of fossil fuels has emerged as a significant factor in the release of carbon emissions.

In conclusion, the results of the study demonstrate a noteworthy impact of population size, energy consumption, and GDP per capita on carbon dioxide emissions. The observed correlation between population size and CO₂ emissions indicates that an increase in population is associated with a corresponding rise in resource and energy consumption, resulting in elevated levels of carbon emissions. The presence of a positive connection between energy consumption and CO₂ emissions highlights the significant impact of energy use on environmental deterioration. This emphasizes the necessity of implementing sustainable energy policies to address and reduce emissions.

In contrast, the observed negative association between GDP per capita and CO₂ emissions suggests the possibility of a decoupling of economic development from carbon emissions. This trend may be suggestive of breakthroughs in cleaner technology, implementation of energy-saving measures, and increased environmental consciousness. However, further research is necessary to fully understand the specific processes that facilitate this trend and to develop policy actions that might effectively enhance this association.

Further research is required to gain more comprehensive knowledge of the intricate dynamics among these variables. This entails considering supplementary elements, such as improvements in technology, legislative frameworks, industrial structures, and socio-economic issues. The examination of how technical innovation and developments in renewable energy sources might serve as a mediator in the interplay of economic growth, population dynamics, energy consumption, and CO₂ emissions is of utmost importance.

In addition, conducting an evaluation of the efficacy of policy initiatives, such as the implementation of carbon pricing systems, incentives for renewable energy, and policies aimed at population control, would yield significant and informative findings. An examination of several nations or areas, taking into account their distinct socio-economic circumstances and policy landscapes, has the potential to enhance comprehension of the intricate interplay between numerous elements that influence environmental results.

By integrating a longitudinal framework to encompass temporal variations and utilizing advanced econometric methodologies, such as dynamic panel data models or structural equation modeling, the study may be improved in terms of its reliability and validity. Furthermore, it is important to engage in multidisciplinary research that fosters collaboration among economists, environmental scientists, policymakers, and technologists. This approach is crucial in order to formulate comprehensive policies that effectively reconcile the objectives of economic expansion and environmental preservation.

By examining these study directions, we may make valuable contributions to the development of well-informed policies and strategies aimed at efficiently managing population growth, energy consumption, and economic development in order to minimize CO₂ emissions and solve the challenges posed by climate change. Ultimately, this will facilitate the establishment of a more sustainable and mutually beneficial relationship between society and the environment.