1. Introduction
The issue of environmental degradation has emerged as a pervasive global concern [
1,
2]. Consequently, authors have expanded their investigations into environmental studies to mitigate the adverse repercussions and safeguard the integrity of the planet’s ecosystem. In the context of national priorities, the fundamental goal of any nation is to enhance its economic progress (GDP) rate to foster social welfare. In this scenario, the occurrence of swift GDP can lead to the reduction of resources and the exacerbation of the environment [
3]. Sustainable GDP in developing nations improves society. Trade, development, foreign direct investment (FDI), and natural resource (NAT) development have been used to achieve this goal. Production increases energy use and CO
2 emissions (CO
2em). GDP must not harm future generations [
4,
5].
It is evident that emerging economies such as the E-7 nations (Brazil, India, China, Indonesia, Mexico, Russia, and Turkey) still have a considerable distance to traverse to achieve sustainable development [
6]. The rapid economic development of the E-7 countries has positioned them as significant contributors to global decision-making processes. The countries mentioned in the study conducted by [
7] represent a significant proportion, exceeding 42% of the global fossil fuel consumption, surpassing the consumption of the G7 countries. Furthermore, as stated in the “Global CO
2 Emissions Report 2019”, developed economies experienced an average GDP rate of 1.7% in 2019. However, it is noteworthy that there was a notable decline of 3.2% in total energy-related CO
2 emissions during the same period. Nevertheless, it is worth noting that the electric power sector has played a significant role in the reduction of emissions, accounting for approximately 36% of energy-related emissions in developed economies. In contrast, in E7 countries, this sector is responsible for more than 47% of CO
2 emissions. Consequently, the E7 nations are assuming a progressively significant role in the global energy market and climate change, encompassing both carbon dioxide emissions and energy consumption. According to [
8], the present predicament faced by E7 countries revolves around the identification of dependable and cost-effective energy alternatives to supplant fossil fuel-based energy sources, all the while mitigating the release of greenhouse gas emissions. Accordingly, emission-mitigation measures must be implemented globally considering the alarming rate of increase in global CO
2 emissions [
9,
10]. More crucially, it has been well postulated in the literature [
11] that greening certain macroeconomic measures can help to decouple CO
2 emissions from economic growth. Purchasing environmentally friendly supplies is crucial in this context. Ref. [
12] suggest that any investment associated with equity should be classified as a green investment (GF). The flow of capital is directed, and establishments are encouraged to actively invest in green projects, which are essential for environmental protection, resource conservation, and economic development [
13,
14].
Environmental degradation is largely attributed to GDP [
15,
16]. However, conventional wisdom holds that early GDP harms the environment but later restores it [
17]. The impact of GDP on climate change has been found to be predominantly negative. However, there are several strategies that can be implemented to mitigate this impact [
18,
19,
20,
21]. The paramount consideration lies in the necessity for both individuals and society to collaboratively undertake concerted endeavors to alleviate these effects [
22]. The significance of a well-educated society should be acknowledged considering its proactive endeavors to safeguard the environment [
23]. Education plays a substantial role in fostering economic development through diverse channels. In broad terms, a society that possesses a high level of education is better equipped to confront a wide range of challenges [
24]. Education is society’s most prized tool for learning new skills and adapting to new technologies [
25]. Several researchers examined education’s impact on the environment. Ref. [
26] found that education is the environmental connection threshold. Secondary education improves environmental quality empirically. Ref. [
27] proposed an education level at which nations start to see a significant positive impact on CO
2em. Thus, education boosts GDP, technological innovation, and production energy source discovery.
The existing body of research examines the correlation between FDI and GDP, the association between renewable energy (RE) consumption and GDP, and the causal connection among GDP, RE consumption, green investments, and CO
2 emissions. Nevertheless, the existing body of research on the moderating role of higher education is relatively scarce. E7 countries exhibit a significant correlation with respect to energy demand, CO
2 emissions, and population levels in comparison to other countries globally. However, there is a scarcity of research studies addressing the current situation in these countries. At present, there exists a dearth of comprehensive empirical studies examining the effects of GDP, RE, FDI, and green investments on carbon emissions, while also considering the controlling influence on education within the E-7 countries. Hence, the present study aims to enhance the analysis by incorporating crucial factors pertaining to interconnectivity through the utilization of an E-7 panel dataset spanning the years 2000 to 2021. For analysis, this work uses the MMQR model. MMQR panel estimation examines the relationship concerning variables across multiple quantiles. The technique by Roger Koenker (2001) [
28] is generally used to approximate the linkages between several factors at different quantiles. Quantile regression is a statistical method that is resistant to the influence of outliers and generates effective estimates for datasets with heavy tails.
This research yields three primary contributions. Firstly, it employs robust estimators, which are analytically appropriate, in its analysis of panel data. Furthermore, the independent variable “portfolio” exhibits variation that differs from that observed in prior studies. Moreover, a clearly defined experimental model is utilized. This study utilizes up-to-date data to obtain empirical findings, with the aim of helping governments and policymakers in the formulation of effective public policies in E-7 countries. Additionally, this measure will contribute to the enhancement of GDP in these economies.
The structure of this work is as follows: after the introduction, the literature review is in
Section 2, while the data and methodology are in
Section 3. The results and discussion are in
Section 4, and the conclusion is in
Section 5.
3. Data and Methodology
Incorporating the moderating effect of education in the E-7 countries, the underlying research work examines the relationship between economic growth, renewable energy consumption, green investments, foreign direct investment, and CO
2em. The E-7 sample shows significant CO
2em, and some of these nations are major carbon emitters, which is why the E-7 economies were chosen for investigation. Because there were available data from 2000 to 2021, we used balanced panel data. Economic growth (estimated as GDP per capita), renewable energy consumption (renewable energy consumption as % of total final energy consumption), green investments (a composite index via principal component analysis using the variables of number of patents, energy efficiency, and renewable energy), and foreign direct investment (FDI) are all listed as independent variables. This work follows the work of [
86] to construct an index of GF. As a dependent variable, CO
2 emissions are included and are expressed as tons of CO
2 per capita.
In the analysis, the role of education is used as a moderating variable. Researchers have used a variety of proxies to estimate economic growth based on prior research. Several well-known studies, including [
87,
88], contend that while many factors influence a nation’s economic growth, GDP per capita can be regarded as a highly reliable indicator. Studies such as [
89,
90,
91] provided guidance for the choice of additional variables and the use of proxy measures. The information for the proposed variables was gathered from several reliable sources. For example, the World Bank’s database was chosen to gather information on FDI information, government spending on education, and GDP per capita. Since the dataset on education contains missing observations, we must first identify these values before we can use a statistical model. These missing values in the dataset are located using the confidence interval method. This method examines each value as it moves from the first to the last element in a sequential order and identifies the observations that are missing. Then, using the mean and standard error of the available observations, it determines the confidence interval using the confidence interval technique. The median of the confidence interval is then used to replace the missing values. The data, their units and sources are mentioned in
Table 1.
Methodology
The interdependence among cross-sectional units has garnered significant attention in contemporary literature, resulting in numerous empirical investigations. The reason for this is that residuals exhibit a lack of independence within the actual context. Therefore, interdependence is inherent. The regional economies that are in the E-7 exhibit cross-border interconnections across various domains, including political, economic, social, environmental, and financial spheres. The correlation implies that any empirical inquiry conducted on these economies must consider the presence of cross-sectional dependence (CD). Equation (1) is proposed for the purpose of investigating the CD in the data.
The correlation between the parameters is represented by
ρN over the time interval T. Furthermore, an inquiry into the heterogeneity present in the slope coefficient has been carried out using the Pesaran and Yamagata (2008) test, which is a revised variant of Swamy’s (1970) test [
93]. Equation (2) has been proposed in this context.
In the context being discussed, N is used to represent the cross-sectional units, while K is utilized to denote the explanatory variables. Upon conducting an examination of the heterogeneity present in the slope coefficients, analysis has been carried out to determine the order of integration between the variables using both the cross-sectionally augmented IPS (CIPS) test and the cointegration augmented Dickey–Fuller (CADF) test. Nonetheless, it is imperative for researchers to give considerable attention to the matter of outcomes during regression estimation.
Hence, the present research has utilized the stationarity test, wherein the cointegration augmented Dickey–Fuller (CADF) test has been elucidated in Equation (3) of this study.
The lagged parameter denoted as
Yt−1 and the initial difference of
Yt−1 is represented by Δ
Yt−1. The computation of CIPS statistics involves the determination of the mean of CADF, which is elucidated in the following equation:
Subsequently, an examination of the cointegrated relationship among the specified factors, namely, renewable energy, GDP, education, and green investments was carried out using Westerlund’s cointegration test. After confirming the cointegration, this work moves forward to apply a novel nonlinear econometric method, namely, the Method of Moments Quantile Regression (MMQR) by [
94]. Traditional linear econometric methods have only focused on modelling the mean of panel data, rather than the conditional distribution [
95]. In contrast, MMQR panel estimation examines the relationship concerning variables across multiple quantiles. The technique by Roger Koenker (2005) [
96] is generally used to approximate the linkages between several factors at different quantiles. Quantile regression is a statistical method that is resistant to the influence of outliers and generates effective estimates for datasets with heavy tails. According to [
97], the method maintains consistency even when multicollinearity is present. However, it should be noted that the quantile regression model exhibits a limitation in its ability to ensure noncrossing outcomes for a multitude of percentiles, which may lead to an inaccurate representation of the response distribution. Considering the context, it is recommended to utilize MMQR owing to a multitude of factors. The model yields consistent results even in the presence of unobserved endogeneity and heterogeneity across the cross-sections. The MMQR methodology allows for the conditional and heterogeneous influence of ecological footprint determinants to impact the distribution’s quantiles. This approach proves advantageous in cases where explanatory variables exhibit high correlation and endogenous behavior.
This method works for high-kurtosis nonlinear datasets. It captures data dynamics endogenously, making it better than other nonlinear modelling methods [
98]. Since parameters depend on response variable location, MMQR allows for asymmetric variable location. Partial-parametric modelling structures like MMQR are ideal for dealing with asymmetry, heterogeneity, and endogeneity producing estimates across numerous quantiles [
99]. The amended location-scale definition for conditional quantiles Q(X) is
The probability (p) can be expressed as
P(
δi + U
itγ > 0) = 1, where (
β,
α,
δ,
γ) are the assessed factors (
βi,
δi),
i = 1, …N, that confirm the fixed effects of individual
i. Here,
U represents a chosen j-vector element of
X that accounts for difference transformation in the equation, denoted as Ul = Ul(X), l = 1, …….j. Furthermore, it can be observed that
Xit denotes an equivalent distribution for a given individual at a distinct point in time (T). Assuming identical distribution at an individual level (i) for time (T), Zit is impertinent to
Xit, Machado and Silva (2019) [
94] instants conditions. The model can be expressed in its quantile in the following manner:
The variables EDU, FDI, GDP, GF, and RE are represented as
Xit, while Q
τ(
τ/X) denotes the dependent factor
Yit, which is the CO
2, conditioned on fundamental quantiles and placed as the independent factor. The notation
βi(τ) =
βi +
δi
p(
τ) is used to denote the quantile τi for an individual. The parameter attains a fixed state and exhibits significant heterogeneous effects, thereby enabling the quantile model τth to manifest through
q(
τ) derived from the linearity problem.
The check function, denoted by the equation above, is
4. Results and Discussion
This work estimates the panel data of E-7 nations. The descriptive statistics are presented in
Table 2.
Table 2 shows that GDP has the highest mean and FDI has the lowest mean. CO
2em has a maximum value of 11.89 and a minimum value of 0.89. RE has a maximum value of 47.11 and a minimum value of 3.18. The graphical form of the descriptive statistics is in
Figure 2.
The next step is to find out the cross-sectional dependence (CD) in the panel data of the E-7 nations. The CD test provides the data description and ensures the applicability of subsequent tests.
Table 3 provides the results, and it shows that all the variables are having CD at the 1% level.
The next step is to know the order of integration of the variables. Some variables may be integrated at level, and some may be at first difference. For this purpose, this work applies the CIPS and CADF unit tests. These tests are better than the first unit root tests of [
100].
Table 4 shows the findings of both unit root tests.
Table 4 shows that all the variables (CO
2, GDP, RE, GF, FDI, and EDU) are cointegrated at first difference. The next step is to know the slope property of the panel data.
Table 5 shows that the panel data are heterogeneous.
Before the long run analysis, it is essential to conduct a cointegration test. For this purpose, this work uses Westerlund’s test [
101]. This test provides robust results while considering the CD in the panel data.
Table 6 shows that the panel data are cointegrated because group and panel values are significant at 1% level. The coefficient values in the panel data further make it possible to know the long-run coefficient values. For the long-run analysis, this work applies the MMQR approach. This method provides the impacts of independent variables on the dependent factor along different quantiles. Next the MMQR results are in
Table 7.
At all quantiles of CO
2 emissions, the estimated coefficient of GDP exhibits a significant positive relationship. The observed trend indicates a positive correlation between the impact of GDP and CO
2em, whereby a higher quantile level corresponds to a higher impact of GDP. Specifically, when CO
2em is situated at a higher quantile level, the effect of GDP is also observed to be higher. This finding aligns with the research conducted by [
67,
102,
103,
104]. The subtext of this statement is that the E-7 nations have experienced positive effects on their economy’s core sectors, including farming, manufacturing, and transportation. Another potential factor to consider is that GDP may stimulate economic activity through the promotion of investment, purchasing, consumption, and energy use, thus leading to a rise in pollution levels [
105].
The impact of RE is negative at all quantiles of CO
2em. This impact is lowering from the 25th to the 90th quantiles. RE is mostly generated from wind, hydro, and solar energy. These energies do not consume fossils and, therefore, do not harm the climate. These results are in line with the findings of [
106].
As far as the impact of GF on CO2em is concerned, our analysis shows that GF is environmentally friendly. This is not an alarming situation for E7 countries because the GF has been able to lower environmental pollution, and it is not a contributing factor towards CO2em. This work constructed the GF as an index composed of the three factors of energy efficiency, RE, and number of patents in E-7 nations. This implies that the allocation of resources towards environmentally friendly initiatives has played a role in reducing CO2 emissions, thereby enhancing the overall climate conditions inside the nation.
The result also suggests that the E-7’s GF policies are aligned with its objective of transitioning towards a low-carbon economy. An alternative hypothesis posits that the energy expenses incurred by industrial and other enterprises in the E-7 countries could be substantial. According to the research conducted by [
107], companies exhibit reluctance towards adopting green technologies, such as green energy, when the energy costs in a particular country are relatively cheap. Entities that incur higher energy costs place a greater emphasis on considering the impact of green initiatives in their decision-making processes, in contrast to entities with lower energy expenditures. The research findings support the legitimacy theory, which argues that corporations should adhere to policies, regulations, and conventions that contribute to environmental sustainability, as evidenced by the negative correlation between GF and CO
2 emissions. Based on this disclosure, it may be inferred that businesses in the E-7 countries adhered to this principle by allocating resources towards the adoption of ecologically sustainable energy sources, machinery, and technology, among other relevant measures. The findings of [
108,
109,
110] support the conclusions of this study. However, the findings of [
111,
112] differ from those mentioned above.
FDI is showing negative signs in all quantiles. This impact is continuously increasing. This means that FDI is suitable to lower the CO
2em in E-7 countries. FDI enables new opportunities for the other developed nations to invest in host countries. It further makes it possible to import efficient technologies from other developed nations into the host countries. These technologies contribute to the lowering of the environmental pollution. The role of education is also friendly in all quantiles. This means that education expenditures are lowering CO
2em in all quantiles, but this impact is lowering while moving to higher quantiles. Education expenditures create opportunities to educate people and to spread environmental awareness. This awareness further encourages citizens to adopt sustainable ways of life. This result is in line with the findings of [
49].
Robustness Check
To check the validity of the MMQR results, this work applies the CCEMG methodology and takes environmental technologies (ENT) as a proxy of green investment. ENT are mainly crafted to deal with environmental pollution and are an authentic variable to measure green investments [
113]. These data have been obtained from OECD. This method is efficient in providing robust results while incorporating the CD in the panel data.
Table 8 shows that RE, GDP, FDI, and EDU impact negatively. This means that these factors are environmentally friendly in the E-7 nations. Green investments are increasing CO
2em. These results are in line with the MMQR results.
5. Conclusions and Policy Suggestions
This research investigates the relationship between carbon dioxide (CO2) emissions, economic growth, renewable energy consumption (RE), green investments (GF), and foreign direct investment (FDI) in the context of higher education in the E-7 countries from 2000 to 2021. The stationarity of the data was assessed through the utilization of three distinct unit root tests. The results of the cointegration analysis indicate the existence of a cointegrating relationship among the variables. The MMQR model was utilized to examine the long-term dynamic relationships among the variables. The findings of the MMQR study indicate that there exists a negative relationship between economic growth, renewable energy (RE), foreign direct investment (FDI), and education (EDU) on the reduction of carbon dioxide (CO2) emissions over an extended period. However, it is worth noting that green investments have the potential to contribute to an increase in CO2 emissions. The findings of this study provide several significant recommendations for policymakers and governments to address the reduction of CO2 emissions while simultaneously promoting sustainable economic growth. Initially, it is imperative for the E-7 economies to allocate a significant proportion of their financial resources towards the advancement of education. There is a need to expand vocational institutions and other educational facilities to accommodate the growing population. In addition to this, it is imperative for governments to give precedence to the establishment of research and development institutions. Through this approach, a multitude of individuals possessing exceptional qualifications will ultimately devise a resolution aimed at mitigating carbon dioxide emissions, facilitated by advancements in technology.
It is important to encourage the use of renewable energy sources like hydrogen, biofuel, biomass, solar, wind, and others as a means of reducing or eliminating our excessive reliance on fossil fuels for economic purposes. Economic incentives that support green services can be encouraged, such as tax waivers or discounted tax rates on goods and services. To secure a sufficient transition to 100% energy from renewable sources, as envisaged by the bulk of recent environmental treaties, the governments should specifically support renewable energy.
Similarly, encouraging investment in innovations, research, and development helps sustain the carbon-mitigating roles of green technology. The governments ought to encourage financial institutions that finance green projects and permit the private sector to participate more in these initiatives. It is difficult to overstate the importance of education in endogenizing advances in technology. Therefore, the governments of the different E7 economies ought to focus more on the education system, especially by reforming and creating curricula that foster the development of creative ideas and skilled labor.
Furthermore, collaboration between the government and policymakers should be undertaken to enact diverse taxation policies with the aim of regulating CO
2 emissions within the environment. The achievement of environmental sustainability can be attained through the expansion of renewable energy sources, as opposed to non-renewable sources, in both product manufacturing and power generation [
114]. E-7 countries should consider implementing policies centered around a system of rewards and penalties to address the issue of environmental degradation. It is imperative for the governments of these nations to acknowledge and incentivize businesses and industries that comply with governmental regulations and employ environmentally sustainable energy sources to meet their production requirements. To mitigate the rise in carbon dioxide emissions, it is recommended that the governments of the E-7 nations take measures to promote the adoption of green investments. Governments must prioritize their efforts to ensure that green financing policies are able to complement environmental welfare policies and green growth policies. Finally, it is imperative for the E-7 nations to optimize the utilization of FDI within their respective economies. To ensure long-term economic sustainability, it is imperative for the governments of the E-7 countries to implement stringent regulatory measures to monitor and control the activities of multinational corporations within their jurisdictions. This approach aims to mitigate environmental degradation and foster the advancement of sustainable development.
Alongside these contributions, this work has some limitations that should be addressed by upcoming research. This work adopts an index of renewable energy, patents, and energy efficiency. Future research can incorporate other proxies of green finance with the application of other robust methodologies of CS-ARDL in different regions. Moreover, other important variables such as institutional quality, financial development, and economic policy uncertainty can be included in the model.