Environmental Sustainability in Emerging Economies: The Impact of Natural Resource Rents, Energy Efficiency, and Economic Growth via Quantile Regression Analysis

Abstract

Improving environmental quality is essential for achieving sustainable economic development when nations pursue growth. Although previous studies looked into different factors of sustainability, the precise effects of natural resource rents as well as renewable energy on CO₂ emissions are yet to be studied in depth. This dissertation attempts to fill the gap by looking at the relationship between economic growth, natural resource rents, renewable energy, and the level of financial development with the environmental quality in eleven regions of emerged and developing economies over the time period of 1990 to 2022. The findings from the Pedroni cointegration analysis reveal a long-run association among financial development, renewable energy, natural resource rents, economic growth, and carbon emissions. Further analysis using the method of moments quantile regression (MMQREG) indicates that renewable energy and natural resource rents significantly reduce CO₂ emissions, particularly at higher quantiles, enhancing environmental quality. Conversely, financial development exacerbates CO₂ emissions, negatively affecting environmental sustainability. Economic growth demonstrates a nonsignificant negative relationship with carbon emissions. The study highlights the critical contributions of renewable energy and natural resource rents to improving environmental quality, while emphasizing the adverse environmental effects of financial development. Policymakers are encouraged to prioritize investments in renewable energy and the effective management of natural resources to mitigate carbon emissions and achieve sustainability in these economies.

1. Introduction

The dynamic relationship between economic growth and environmental sustainability increasingly captured scholarly attention, particularly in emerging economies where rapid development often intensifies energy demand and carbon emissions [1,2,3,4]. The environmental Kuznets curve (EKC) theory posits that environmental degradation may initially worsen with economic expansion before eventually improving past a certain income threshold [5,6,7]. However, much of the existing literature remains largely descriptive, failing to critically unpack the asymmetric and context-dependent relationships among economic growth, financial development, energy consumption, and other key factors that drive environmental outcomes.

Recent studies have begun to challenge this one-dimensional view. For example, cleaner energy options and robust financial systems have been identified as potential mitigators of the adverse environmental impacts associated with economic expansion [8,9,10]. Empirical evidence suggests that renewable energy adoption can significantly reduce carbon emissions in developing regions [11,12]. At the same time, financial development plays a dual role: while early-stage financial expansion may spur emissions through industrial growth, more mature financial systems tend to promote green investments and sustainable technologies [13,14]. Moreover, the structure of energy consumption—specifically, the balance between non-renewable and renewable sources—critically shapes environmental outcomes [13,15], and the overall impact of these factors can vary substantially across regions [2,16,17,18]. Additionally, governance and institutional quality further influence these dynamics by reinforcing the positive effects of financial development on environmental quality [19,20].

In addition to these well-studied factors, two other critical elements warrant attention. First, natural resource rents—which reflect the economic benefits derived from natural resource extraction—can significantly influence both economic growth and environmental sustainability. In emerging economies, reliance on natural resources may boost GDP per capita, yet, if these rents are not managed sustainably, they can exacerbate environmental degradation. Second, energy efficiency is a vital mechanism for reducing the environmental impact of economic activities. By lowering the amount of energy required per unit of output, energy efficiency measures can mitigate carbon emissions even as economies expand.

Although these insights are valuable, a critical gap remains in the literature. Researchers need to systematically investigate how several factors interact in emerging economies. These factors include economic growth, natural resource rents, energy efficiency, financial development, and energy consumption. These factors appear to interact asymmetrically. Thus, addressing this gap becomes essential, as it will help develop targeted environmental policies that reconcile economic development with sustainability goals.

In light of this, the objectives of the present study are to:

• Critically assess the asymmetric effects of economic growth on carbon emissions in a sample of eleven emerging economies.
• Evaluate the moderating roles of financial development, energy consumption (including the balance between renewable and non-renewable sources), natural resource rents, and energy efficiency in shaping the environmental impact of economic growth.
• Derive context-specific policy recommendations that can effectively harmonize economic advancement with environmental sustainability.

These objectives give rise to the following research questions:

Q₁: How does economic growth asymmetrically influence carbon emissions in the selected emerging economies?

Q₂: In what ways do financial development, energy consumption patterns, natural resource rents, and energy efficiency modify the impact of economic growth on environmental quality?

Q₃: What targeted policy interventions can be formulated to balance economic development with environmental sustainability in these nations?

Following this background, this paper is structured into four main sections. The Section 2 reviews the empirical literature relevant to the study. In the Section 3, the methodology and analytical tools are described. The Section 4 focuses on analyzing and discussing the findings in depth. Lastly, the paper is summarized in the Section 5.

2. Empirical Literature Review

2.1. Economic Growth–Carbon Emissions Dynamics

The relationship between economic development and environmental degradation garnered substantial scholarly attention, with many studies focusing on how economic growth influences carbon emissions across various regions. A central theoretical framework in this debate is the environmental Kuznets curve (EKC) hypothesis, which posits an inverted U-shaped relationship whereby emissions initially increase with economic growth and later decline as higher income levels spur the adoption of cleaner technologies and sustainable practices. However, despite its widespread use, the EKC hypothesis remains contested due to mixed empirical evidence and context-specific variations.

For instance, ref. [21] provides evidence from the economies of the Union for the Mediterranean, showing that advancing economic development is associated with increased investment in eco-friendly technologies. While this finding supports the notion of a turning point in the growth–emissions nexus, its reliance on regional-specific data raises questions about the broader applicability of the results. Similarly, ref. [1] examined the Chinese context, highlighting the pivotal role of green innovation in mitigating the adverse environmental impacts of economic expansion. Although their study suggests that innovative and sustainable solutions can reverse initial emission increases, the broad categorization of “green innovation” may obscure the diverse technological pathways and policy contexts that contribute to these outcomes.

Ref. [3] extends the discussion by exploring the interaction between economic growth, urbanization, and foreign trade in China. Their work reveals that carbon emissions tend to increase in areas characterized by intense urbanization and industrial activity. However, the study remains primarily descriptive and does not fully interrogate the causal mechanisms through which these interactions occur. In contrast, research b [22,23] distinguishes between aggressive and moderate growth targets, arguing that unsustainable practices driven by rapid expansion hinder emission reduction, whereas moderated economic goals may foster innovation and better environmental outcomes. Although these findings offer valuable insights, they often overlook the potential moderating effects of policy interventions and institutional quality.

Energy consumption emerges as another critical conduit linking economic growth and emissions. Ref. [24], using spline analysis in the United States, finds that the relationship between growth and emissions is comparatively weak in developed economies—a result that underscores the distinct energy consumption patterns between developed and developing nations. Supporting this perspective, ref. [25] stresses the importance of balancing economic growth with environmental objectives. However, his analysis tends to simplify the complex, nonlinear interactions between sector-specific energy use and overall economic expansion. Moreover, demographic factors, as evidenced by [26] in their study of population aging, suggest that shifting demographic profiles can weaken the traditional growth–emissions linkage. This observation introduces a critical dimension that warrants further investigation to assess its consistency across different economic contexts.

At a more granular regional scale, ref. [27] demonstrates that economic growth does not uniformly lead to increased emissions across Chinese cities; rather, local economic structures and policy frameworks significantly shape emission patterns. In a similar vein, ref. [14] illustrates in the Nigerian context how financial development can modulate the long-term relationship between growth and carbon emissions. While these studies advance the discourse by integrating financial and institutional factors, they also highlight a need for more in-depth analyses that can elucidate the underlying causal pathways.

Additional studies reinforce the multifaceted nature of this dynamic. Ref. [25] underscores the influence of industrial shifts and energy intensity, while [16] emphasizes the critical role of energy efficiency improvements in achieving sustainable economic growth. Nonetheless, both studies tend to rely on cross-sectional data, which limits their ability to capture dynamic temporal changes in the growth–emissions relationship.

Furthermore, research from other regions adds to the complexity of the debate. Refs. [28,29], in an African meta-analysis, document that unsustainable industrial practices and weak environmental regulations contribute to rising carbon footprints. These studies, however, largely stop short of critically assessing how targeted policy interventions might reverse these trends. Ref. [30] documents a similar barrier in Indonesia, where energy consumption and emissions impede economic progress, thereby emphasizing the urgency for sustainable energy policies. However, similar to several other studies in this review, their work remains predominantly correlational, leaving causal mechanisms insufficiently explored.

The role of international trade further complicates the narrative. Ref. [31] notes that while trade and growth initially contribute to higher emissions, technological progress in more advanced stages may reverse this trend, which is in line with the EKC hypothesis. In South Africa, ref. [32] identifies a robust long-term link between economic growth and emissions, with trade openness and foreign direct investment playing critical roles. However, both studies tend to underemphasize the influence of domestic policy frameworks that might mediate these relationships, suggesting an area ripe for further research.

In summary, while the existing literature provides a broad overview of the factors linking economic growth to carbon emissions—ranging from energy consumption and demographic shifts to international trade and financial development—a number of critical gaps remain. Many studies adopt descriptive approaches that do not sufficiently interrogate the causal mechanisms at play or the context-specific conditions that drive these dynamics. A more critical examination of these factors is essential to inform the design of effective policies that can harmonize economic advancement with environmental sustainability.

2.2. Financial Development–Carbon Emissions Nexus

The literature on financial development’s impact on environmental health, particularly carbon emissions, reveals a complex and context-dependent relationship. Several studies indicate that the effects of financial development can be either beneficial or detrimental, contingent upon the underlying financial infrastructure, regulatory environment, and investment allocation.

On one hand, enhanced financial development can improve access to capital, enabling investments in eco-friendly technologies and renewable energy projects that reduce carbon emissions. For example, ref. [33] documents that in BRICS nations, a more developed financial sector supports energy-efficient and renewable energy projects, thereby contributing to a smaller carbon footprint. Similarly, ref. [34] argues that in developing countries, financial growth facilitates green financing initiatives, promoting long-term reductions in emissions.

Conversely, financial development may also lead to an increase in carbon emissions when capital is channeled toward sectors with high carbon intensity. Ref. [35] observes that in oil-driven economies such as Nigeria and Iran, the expansion of fossil fuel-dependent industries significantly raises emissions. In a similar vein, ref. [4] highlights that in many developing nations, financial development is often correlated with rising carbon emissions, particularly in environments characterized by limited regulatory control where energy-intensive industries tend to attract the most capital.

The relationship between financial development and carbon emissions is thus far from uniform. Ref. [36] emphasizes that when financial growth is paired with robust environmental regulations and targeted incentives for renewable energy investments, it can help mitigate carbon emissions. In contrast, ref. [35] illustrates that in economies prioritizing rapid economic expansion over environmental concerns, the absence of such regulatory measures often leads to increased emissions. Moreover, ref. [37] warns that a lack of green financial products and sustainable investment practices may exacerbate environmental degradation, especially in regions heavily reliant on fossil fuels.

In summary, the nexus between financial development and carbon emissions is multifaceted. While financial development holds the potential to reduce emissions through the promotion of cleaner technologies and renewable energy, its overall effect is largely dependent on the strength of environmental regulations and the prioritization of sustainability within financial systems.

2.3. Energy Consumption-Carbon Emissions Nexus

At the heart of many studies lies the environmental Kuznets curve (EKC) hypothesis, which posits that while early stages of economic growth may lead to higher emissions, a turning point is eventually reached at which further growth coincides with environmental improvements [6,17,21]. For example, ref. [38] found in Morocco that economic expansion initially exacerbates CO₂ emissions, but with the right mix of innovation and regulation, improved environmental quality can follow. Similarly, refs. [3,23] demonstrate that factors such as urbanization, industrial activity, and foreign trade further complicate this relationship, with regional studies [28,29,30] underscoring the context-specific nature of these dynamics. Complementing these findings, ref. [2] highlighted that increased energy demand driven by economic growth often leads to higher emissions, a pattern further nuanced by the influence of governance and institutional quality [19,20].

The role of financial development in this nexus is equally multifaceted. On the one hand, a robust financial sector can mobilize capital for investments in clean and renewable technologies, thereby reducing the carbon footprint [33,34]. On the other hand, in environments where regulatory frameworks are weak, financial expansion may inadvertently direct investments toward carbon-intensive industries, leading to increased emissions [4,35]. Refs. [35,36] argue that the beneficial impact of financial development on the environment is conditional on strong environmental regulations and incentives, a view further reinforced by [37] in the context of green finance.

Energy consumption is perhaps the most direct driver of carbon emissions, with the type of energy source playing a decisive role. Fossil fuels, such as coal, oil, and natural gas, remain dominant in many developing economies and are closely linked to rising emissions [4,33,35,39]. In contrast, the transition to renewable energy sources, such as solar, wind, and hydropower, offers promising avenues for reducing greenhouse gas emissions [9,40]. However, the pace of this transition is often slow. Refs. [34,35] note that despite incremental increases in renewable energy adoption, many nations continue to grapple with a heavy reliance on fossil fuels. This dynamic is further illustrated by [24] in developed economies and stressed by [25] in emerging contexts. Refs. [27,41,42] additionally shows that local economic structures and policies can lead to significant variations in how energy consumption translates into carbon emissions.

Recent regional studies deepen our understanding of these linkages. Ref. [43], employing a vector error correction model (VECM) in the Western Balkans, found a long-term relationship between energy consumption and carbon emissions, indicating that higher energy use may be a persistent barrier to environmental quality. This finding resonates with the work of [44], whose panel cointegration analysis across eight Asian countries confirms that elevated energy consumption is a key contributor to increased CO₂ levels. In China, research focusing on Sichuan Province by [44,45] reveals that coal remains the dominant source of emissions, underscoring the need for stringent controls on coal use. Complementarily, refs. [16,45] advocate for policies that promote renewable energy and energy efficiency, while [46] emphasizes the impact of urban planning—specifically residential building layouts—on reducing energy demand and associated emissions. Ref. [45] further integrates carbon emissions into a broader water–energy–carbon nexus framework, offering insights into how coordinated urban resource management can facilitate emission reductions.

Technological factors add another layer of complexity to the relationship between energy use and emissions. While advancements in information and communication technology (ICT) can drive economic growth and improve energy efficiency [47], they may also lead to higher energy consumption if not properly managed—a point illustrated by [48] through the application of the STIRPAT model. Predictive models, such as those developed by [25] using the LMDI approach and by [49], reveal a negative relationship between alternative energy consumption and carbon emissions, reinforcing the importance of shifting toward renewable energy sources for long-term sustainability.

Regional analyses also highlight the need for a diversified energy mix. In Azerbaijan, for instance, ref. [50] demonstrates that increasing the share of nuclear and renewable energy can mitigate emissions, whereas [51] points to the continued dominance of fossil fuels as a major challenge. Such findings stress that achieving meaningful emission reductions will require integrated strategies that combine technological innovation, targeted policy interventions, and a balanced mix of energy sources.

In summary, while economic and financial advances offer pathways to innovation and improved energy efficiency, they also risk exacerbating environmental degradation if not supported by robust regulatory frameworks and a strategic shift toward renewable energy. Integrated approaches that embrace technological innovation, sound policy measures, and diversified energy portfolios are thus essential for effectively addressing the multifaceted challenges of climate change.

2.4. Influence of GDP per Capita on Economic Growth and Environmental Quality

GDP per capita is a common measure of economic growth. It also shapes environmental outcomes, especially carbon emissions. The environmental Kuznets curve (EKC) hypothesis suggests a nonlinear relationship. Early economic growth often worsens environmental degradation. However, after reaching a certain income level, cleaner technologies and stronger policies help reduce emissions [37,52].

Empirical evidence in some developing economies supports this view. Ref. [35] shows that countries such as South Korea, Mexico, and Turkey saw rising emissions during early industrial expansion. As these nations grew wealthier, investments in renewable energy and energy-efficient technologies helped lower emissions. Ref. [9] reports similar trends in Southeast Asia, including Vietnam and the Philippines.

However, the EKC does not hold everywhere. In many low- and middle-income countries, financial constraints and weak institutions limit progress. Ref. [33] notes that in Nigeria and Bangladesh, economic growth has not reduced emissions due to heavy reliance on fossil fuels and poor environmental policies. Ref. [37] further emphasizes that ineffective regulations hinder efforts to control emissions.

Critics argue that the link between GDP per capita and carbon emissions is more complex than the EKC suggests. Ref. [34] stresses that economic structure, energy policies, and technological progress all influence this relationship. In economies dependent on non-renewable energy, industrial growth often drives higher emissions, regardless of income. This trend is seen in oil-dependent nations such as Iran and Nigeria, where rising income has not curbed carbon emissions [35].

The role of technology and policy is critical. Ref. [40] shows that countries with strict environmental regulations and investments in green innovation can achieve growth while reducing emissions. In sum, GDP per capita is a key driver of carbon emissions. Yet, its impact varies widely. In some contexts, rising incomes lead to cleaner environments. In others, they exacerbate environmental harm due to ongoing fossil fuel dependence and weak regulations.

2.5. Asymmetry in the Effect of Economic Growth on Carbon Emissions

Recent research reveals a complex and asymmetric link between economic growth and carbon emissions. During periods of expansion, emissions surge sharply due to intensified industrial activity and heavy reliance on fossil fuels [52,53]. In contrast, economic downturns yield only modest reductions. Industries tend to maintain a baseline level of energy use, and governments often relax environmental policies to boost recovery.

Ref. [36] documents this imbalance in developing nations such as Iran, Nigeria, and Pakistan. Their study shows that positive economic shocks lead to large increases in emissions, while negative shocks only produce small decreases. Ref. [4] further emphasizes that in economies dependent on fossil fuels, the carbon footprint of growth far exceeds any gains from slower activity.

Ref. [33] argues that the stage of economic development shapes this asymmetry. In developing countries with outdated, non-renewable energy systems, the positive link between growth and emissions is particularly strong. Even in downturns, energy demand remains high in energy-intensive sectors such as manufacturing, limiting emission reductions [35]. By contrast, advanced economies such as South Korea and Mexico mitigated this effect by investing in renewable energy and improving energy efficiency.

Government policies in emerging markets further exacerbate this imbalance. Ref. [40] reports that during economic booms, governments often prioritize growth over environmental protection, thereby driving up emissions. In downturns, stricter environmental measures are delayed to avoid hindering recovery [34]. Moreover, sectoral analysis shows that while the industrial sector significantly increases emissions during growth, sectors such as transport and services remain less responsive, sustaining emissions even during contractions [4]. Ref. [37] stresses that major shifts in energy use are essential for high-emission countries such as Turkey, Vietnam, and Indonesia.

In summary, the evidence points to a clear asymmetry in how economic growth affects carbon emissions. Growth periods produce significant spikes in emissions, whereas downturns yield only marginal declines. This imbalance arises from structural energy dependencies and policy choices. To address this issue, targeted measures are needed. This includes policies that enforce strict environmental regulations and promote renewable energy development, even during periods of strong economic performance.

2.6. Research Contributions

Although the relationship between economic growth and carbon emissions is well researched, our study addresses key gaps in literature. First, prior studies often treat the growth–emissions nexus as symmetric. In contrast, our findings reveal that emissions react very differently during periods of economic expansion compared to downturns [52,53]. This asymmetry, where positive shocks drive sharp increases in emissions, while negative shocks yield only modest declines, received limited attention despite its policy relevance [4,36].

Second, while many studies examine the individual roles of GDP per capita, financial development, or energy consumption on environmental outcomes, few integrate these factors into a unified analytical framework. By considering these dimensions simultaneously, our paper offers a more nuanced understanding of how economic, financial, and energy factors interact to shape carbon emissions. This integrated perspective is crucial for emerging and developing economies, where diverse structural and policy challenges often coexist [33,35].

Third, the methodological contribution of this paper is significant. Unlike conventional approaches that estimate average effects, we apply the MMQREG technique to capture heterogeneity across the distribution of carbon emissions. This approach reveals differential impacts of economic growth, financial development, and energy consumption at various quantiles. In doing so, it uncovers important patterns that standard mean regression models may overlook, thereby offering insights that are critical for designing targeted and effective environmental policies.

Furthermore, our study advances the debate on the environmental Kuznets curve (EKC) by exploring whether its nonlinear pattern holds uniformly across different emission levels and economic contexts. By dissecting the asymmetric relationship between growth and emissions, we clarify how contextual factors such as energy mix and regulatory frameworks influence the EKC dynamics in emerging markets.

In summary, the novelty of our research lies in three key areas: (1) highlighting the asymmetric effects of economic growth on carbon emissions; (2) integrating multiple determinants within a single analytical framework; and (3) employing a novel MMQREG methodology to expose heterogeneous effects across the emission spectrum. These contributions not only deepen our theoretical understanding of the growth–environment nexus, but also offer practical insights for policymakers aiming to balance economic development with environmental sustainability.

3. Methodological Issues

The study employs annual data for 11 countries—Egypt, Bangladesh, Korea Republic, Mexico, Iran, Nigeria, Vietnam, Indonesia, Turkey, Pakistan, Indonesia, and the Philippine—culled from the world bank data indicator (WDI). These countries were chosen due to their representation of emerging and developing economies with notable dependence on natural resources, different degrees of financial development, and distinct energy consumption, thereby providing an excellent opportunity to analyze the asymmetric impacts of the explanatory variables on environmental quality. Moreover, the N-11 countries together with the BRICS countries, according to the Goldman Sachs investment bank, are believed to be the greatest economies of the 21st century. Therefore, it is essential to understand environmental sustainability in these emerging economies. This is so because most emerging economies, such as China and India, among others, achieved development of their economy at the expense of degrading the environment. The data for this study span from 1990 to 2022 for carbon emission (CO₂) serving as the explained variable, while financial development (proxied with domestic credit to private sectors), renewable energy (REN), total natural resources rent (TNR), and gross domestic product (GDP) serve as the explanatory variables in re-examining the asymmetric effect of economic growth on environmental quality. The selection of GDP, FD, REN, and TNR as exogenous variables is based on strong economic theory and empirical evidence. GDP tests the environmental Kuznets curve (EKC) hypothesis, evaluating whether economic growth in emerging economies leads to lower emissions over time or continued environmental degradation. FD captures the dual effect of financial development, which can either increase emissions by funding industrial expansion [54] or reduce emissions by promoting green investments [34]. REN assesses the effectiveness of renewable energy adoption in reducing emissions, particularly at higher pollution levels [9], while TNR tests the resource curse hypothesis, determining whether resource wealth worsens emissions [14] or supports sustainable development [55]. These variables allow for a comprehensive analysis of sustainability transitions, helping policymakers understand the key economic drivers of environmental quality in emerging economies.

We aim to reinvestigate the elasticities among financial development (proxied with domestic credit to private sectors), renewable energy, total natural resources rent, GDP, and CO₂ within the environmental Kuznets curve (EKC) theory. However, there is an existing indirect relationship between the independent variables vector and environmental quality theoretically pertaining to the EKC literature [15,20]. Hence, carbon dioxide (CO₂)—the dependent variable—is regressed as a function of financial development, renewable energy, total natural rent, and GDP as the explanatory variables employed in the study.

This study employs a panel data approach to examine the asymmetric impacts of economic growth, financial development, renewable energy, and natural resource rents on environmental quality across 11 emerging and developing economies from 1990 to 2022. The econometric analysis is structured into multiple steps to ensure robustness and address potential econometric concerns such as cross-sectional dependence, heterogeneity, and endogeneity. The relationship between carbon emissions (CO₂) and its key determinants—financial development (FD), renewable energy consumption (REN), natural resource rents (TNR), and economic growth (GDP)—is modeled within the environmental Kuznets curve (EKC) framework. The general functional form is specified as follows:

$$c{o}_{2it}={\alpha }_0+{\alpha }_{1}f{d}_{it}+{\alpha }_{2}re{n}_{it}+{\alpha }_{2}tn{r}_{it}+{\alpha }_{2}gd{p}_{it}+{\epsilon }_{it}$$

(1)

where i represents the country, t represents the time period, and $\epsilon$ is the error term.

Given the potential for heterogeneous impacts across different levels of carbon emissions, the study employs the maiden method moments quantile regression (MMQREQ) panel estimation technique as proposed [56], which allows for the estimation of the effects of explanatory variables at different points of the CO₂ distribution, thereby capturing nonlinearities and asymmetric responses.

The conditional quantile provides distributional and heterogenous impacts in a different location-scale of the CO₂ as follows:

$${\mathbb{Q}}_{co{2}_{i,t}}\left(\tau |{\alpha }_i,{\epsilon }_{t,}{X}_{it}\right)={\alpha }_{\tau }+{\gamma }_{1a\tau }lf{d}_{it}+{\gamma }_{2a\tau }lre{n}_{it}+{\gamma }_{3a\tau }ltn{r}_{it}+{\gamma }_{4a\tau }gd{p}_{it}+{\epsilon }_{it}\left(\tau \right)$$

(2)

where $c{o}_2$ denotes carbon dioxide, $X_{it}$ is a $k\times 1$ vector of the explanatory variables, ${\mathbb{Q}}_{co{2}_{i,t}}\left(\tau |{\alpha }_i,{\epsilon }_{t,}{X}_{it}\right)$ is the conditional $\tau -th$ quantile of CO₂, given REN, FD, TNR and GDP, ${\alpha }_{\tau }$ is the intercept term of the $\tau -th$ quantile; ${\gamma }_{1a\tau },{\gamma }_{2a\tau },{\gamma }_{3a\tau } \mathit{and} {\gamma }_{4a\tau }$ represents the coefficients for FD, REN, TNR, and GDP at the $\tau -th$ quantile, respectively, while ${\epsilon }_{it}\left(\tau \right)$ is the error term quantile $\left(\tau \right)$ specific error term. It is of note that for a specific quantile $\left(\tau \right)$, the quantile regression model involves minimizing this objective:

$${\mathbb{Q}}_{\tau }\left(\beta \left(\tau \right)\right)={\displaystyle \sum _{i=1}^N{\displaystyle \sum _{t=1}^T{\rho }_{\tau }\left(y_{it}-X_{it}\beta \left(\tau \right)-{\alpha }_i\right)}}$$

where ${\rho }_{\tau }\left(u\right)=u\left(\tau -I\left(\tau <0\right)\right)$; $\tau$ is the quantile of interest (e.g., 0.5, 0.25 and 0.75 for the median, lower, and upper quartiles, respectively). Unlike traditional mean regression models, MMQREG estimates the impact of independent variables at different quantiles (e.g., 25th, 50th, and 75th percentiles), allowing for a more comprehensive understanding of distributional effects. This approach is particularly useful for identifying whether financial development, natural resource rents, and renewable energy have stronger effects at lower or higher levels of carbon emissions.

Before conducting panel MMQREG analysis, the dataset’s cross-sectional dependence, slope heterogeneity, and stationarity tests were conducted. For instance, the cross-sectional dependence focusses on a shock affecting one unit in the panel that may also affect others, while the slope coefficient computed for the entire panel might be inconsistent for the individual panel members, as estimated slope coefficients may not be homogenous (see [57]). Based on this, we first examine our dataset’s cross-sectional dependence and slope heterogeneity before conducting panel data analysis.

The MMQR analysis fails to check the causality between the variables, but only examines them at a certain scale and location. Therefore, we used the panel data causality estimation, put forward by [58]. The Granger causality analysis addresses the unbalanced panel more efficiently and effectively. It also handles the panel data heterogeneity and crosssectional dependence. The MMQR method is crucial in panel data analysis because it presents results in different quantiles [59]. This is essential in examing the asymmetric effects in the different quantiles [60]. The MMQR method is also fundamental because it overcomes heterogeneity and CD problems, that is, it is the second-generation method with the capacity to overcome these issues [61]. This makes the MMQR method superior over other methods of data analysis, such as the first-generation methods that fail to adjust for CD and heterogeneity.

We use the CD test [62], the biased-adjusted LM test by [63], and the $\widetilde{∆}$ cum ${\widetilde{∆}}_{adj}$ by [64] for slope heterogeneity. It should be established that the CD test is robust to the non-normality of errors and structural breaks as it focuses more on the cross-sectional unit of the panel than the times’ dimension. At the same time, the bias-adjusted LM was used as a supplement to the CD test’s result. The CD—Equation (3)—and bias-adjusted CD tests—Equation (4)—are presented thus:

$$CD=\sqrt{{\displaystyle \frac{2T}{N(N-1)}}}\left({\displaystyle \sum _{i=1}^{N-1}{\displaystyle \sum _{j=i+1}^N{\widehat{\rho }}_{ij}}}\right)\sim N(0,1)i,j=1,2,3,\dots ,N$$

(3)

$${LM}^{*}=\sqrt{{\displaystyle \frac{2T}{N(N-1)}}}\left({\displaystyle \sum _{i=1}^{N-1}{\displaystyle \sum _{j=i+1}^N{\widehat{\rho }}_{ij}}}\right){\displaystyle \frac{(T-k){\widehat{\rho }}_{ij}^2-E(T-k){\widehat{\rho }}_{ij}^2}{Var(T-k){\widehat{\rho }}_{ij}^2}}$$

(4)

where ${\widehat{\rho }}_{ij}^2$ is a sample estimate of the pairwise correlation between the OLS residuals.

The slope heterogeneity was tested using the [64] test, which accommodated the presence of cross-sectional dependence, whereas the slope homogeneity test uses the below test statistics:

$$\tilde{S}={\displaystyle \sum _{i=1}^N{\left({\widehat{\beta }}_i-{\tilde{\beta }}_{WFE}\right)}^{\prime }\frac{\overline{X^{\prime }}\overline{X}}{{\tilde{\sigma }}_i^2}}\left({\widehat{\beta }}_i-{\tilde{\beta }}_{WFE}\right)$$

(5)

$$\widehat{∆}=\sqrt{N}{(2k)}^{{\displaystyle \frac{-1}{2}}}(N^{-1}\widehat{S}-k)$$

(6)

$$\widetilde{∆}=\sqrt{N}{(2k)}^{{\displaystyle \frac{-1}{2}}}(N^{-1}\tilde{S}-k).$$

(7)

$\tilde{S}$, $\widehat{∆}$, and $\widetilde{∆}$ are statistical tests, $\overline{X}$ is the independent variable vectors in the mean deviation, and ${\tilde{\beta }}_{WFE}$ captures the weighted fixed effects estimators—the weights are constructed using ${\tilde{\sigma }}_i^2$—and k denotes regressors’ number. The mean and variance bias-adjusted versions of $\widehat{∆}$ and $\widetilde{∆}$ are denoted below:

$${\widehat{∆}}_{adj}=\sqrt{N}{\left({\displaystyle \frac{2k{(T-k-1)}^2(T-s)}{{(T-k-3)}^2(T-k-5)}}\right)}^{{\displaystyle \frac{-1}{2}}}\left(N^{-1}\widehat{S}-{\displaystyle \frac{k(T-k-1)}{T-k-3}}\right)$$

(8)

$${\widetilde{∆}}_{adj}=\sqrt{N}{\left({\displaystyle \frac{2k(T-k-1)}{T+1}}\right)}^{{\displaystyle \frac{-1}{2}}}(N^{-1}\tilde{S}-2k).$$

(9)

4. Result and Discussion

The empirical analysis first explores the properties of the variables utilized using the descriptive statistics tools as summarized in

Table 1. Summary statistics.

Statistics	CO2	FD	GDP	REN	TNR
Mean	9.670	37.821	4.497	30.114	6.175
Median	0.000	27.388	4.985	26.600	2.602
Maximum	174.968	174.968	15.329	88.600	34.779
Minimum	0.000	0.000	−13.127	0.000	0.000
Std. Dev.	33.079	32.173	3.346	26.305	7.820
Skewness	3.461	1.973	−1.120	0.656	1.705
Kurtosis	13.806	6.862	6.836	2.365	5.426
Jarque–Bera	2490.755	460.997	298.433	31.968	264.857
Probability	0.000	0.000	0.000	0.000	0.000
Observations	363	363	363	361	363

Source: Author’s own work (2024).

. The summary table shows that all the variables exhibited a leptokurtic peakedness and positively skewed, except renewable energy (REN), with a platykurtic kurtosis and GDP that skewed negatively, respectively.

Interestingly, all the variables’ Jarque–Bera statistics are significant, hence connoting non-normality of the variables, and since normality is one of the basic assumptions of the OLS estimator, all OLS estimation techniques becomes non-applicable, thereby reaffirming the exigency of the panel quantile regression that accounts for the heterogeneity of carbon emission by quantiles.

Furthermore,

Table 2. VIF results of multi-collinearity.

Variable	VIF	1/VIF
$lfd$	1.25	0.9985
$lren$	1.10	0.9074
$ltnr$	1.28	0.7795
$lgdp$	1.03	0.7723
Mean VIF	1.17

depicts that all the explanatory variables in the model are free from multicollinearity, as shown through the variance inflator factor (VIF), which is less than 10, establishing that the independent variables in the model are not strongly related and hence can be specified together in the same model.

Furthermore, the cross-sectional dependency or otherwise among the countries studied are examined as shown in

Table 3. Cross-sectional dependence.

Variables	LCO2	LFD	LREN	LTNR	GDP
CD	15.09 ***	2.25 **	15.24 ***	13.47 ***	7.16 ***
Abs (Corr)	0.761	0.447	0.459	0.446	0.204

Source: Author’s own work (2024). **, *** indicate that statistics are significant at the 5% and 1% level of significance, respectively. The null hypothesis is no cross-sectional dependence.

:

Going forward, we conducted the panel unit root test since cross-sectional dependency has been established; the CIPS panel unit root test that accommodates cross-sectional dependence and [65] IPS first-generation unit root test was utilized to achieve this. The unit root result is presented in

Table 4. Panel unit root tests.

	CIPS	IPS
	Null: Homogenous Non-Stationary	Null Hypothesis: Unit Root with Individual Process
$lco2$	−4.756 ***b	−9.27166 ***b
$lfd$	−4.291 ***b	−8.9908 ***b
$lren$	−3.659 ***b	−1.8982 **b
$ltnr$	−5.052 ***b	−10.2148588 ***b
$lgdp$	−4.641 ***b	−12.6679 ***b

Source: Author’s own work (2024). Note: ^b denotes stationarity at level and first difference, while *** and ** indicate statistical significance at 1% and 5%, respectively.

.

The CIPS test disclosed that all the variables of interest were stationary after first differencing; the same scenario played out using the IPS first-generation test where all were stationary I(1). Since the study was initially exposed to cross-sectional dependency in the panel data, we align with the CIPS submission of I(1).

The tests for slope homogeneity reveal that the slope is heterogeneous, as is manifest in

Table 5. Slope homogeneity test result.

	Value
Delta tilde ($\widetilde{∆}$)	17.033 ***
Delta tilde adjusted (${\widehat{∆}}_{adj}$)	18.831 ***

Source: Author’s own work (2024). Note: *** indicate that statistics are significant at 1% significance level. The null hypothesis is no cross-sectional dependence.

, as the critical values, as denoted in Pesaran and Yamagata, were less than the test value. Since the results suggest the existence of cross-sectional dependence and slope heterogeneity (see Table 3 and Table 5), an estimation methodology and estimator that is efficient for these two is employed for this study. Furthermore, we examined whether a long-run nexus exists in the empirical model using the Pedroni panel cointegration tests.

Table 6. Panel cointegration test.

	Statistics	p-Value
Modified Phillips-Perron t	−2.5297 ***	0.0057
Phillips-Perron t	−8.5784 ***	0.0000
Augmented Dickey–Fuller t	−8.9330 ***	0.0000

Note: *** indicates that statistics are significant at a 1% significance level. The null hypothesis is no cointegration.

presents the empirical result for the MMQREG analysis for the 11 countries examined after confirming a long-run association (see Table 5) among the variables and establishing that the variables are not normally distributed using the Jarque and Bera normality test. It depicts that a percentage change at the quantile $Q_{0.25} \mathit{and} Q_{0.50}$ in LFD ignites a 1.3508 and 1.1733 increase in CO₂, respectively. Meanwhile, at the 75th and 90th quantiles, the values are also positive but statistically not significant, establishing that the LFD only influences CO₂ in the studied panel at the 25th and 50th quantile. Noticeably, the values from the lesser to higher quantiles have a significant magnitude change, portraying that the positive impact of LFD tends to reduce in higher quantiles. This is possible as this deals with financial resources provided to the private sector by financial corporations, through loans, purchases of nonequity securities, and trade securities, hence, domestic credit to the private sector moderates energy consumption to increase carbon emissions. The results do not confirm the existence of the EKC hypothesis but confirm that domestic credit to the private sector increases carbon emissions (CO₂). Additionally, in relative to environmental advantages, domestic credit to the private sector contributes towards employment opportunities, economic expansion, and poverty alleviation cum reduction [2,31,66,67].

The results for renewable energy (REN) show a significant negative influence on CO₂ emission from 0.6441 to 1.7366, from lesser (25%) to higher (75% and 90%) quantiles, indicating that the influence of renewable energy develops higher at higher quantiles. Correspondingly, for total natural resources rent (TNR), there exists a significant and negative impact on carbon emission (CO₂), as the magnitude effects steam higher from a lower quantile to a higher quantile. Contrarily, GDP estimates show a negative but insignificant influence on carbon emission, hence contrasting earlier research studies (see [8,15,18]). Furthermore, all the aforementioned factors’ effects are depicted graphically in quantiles (see Figure 1), depicting how the magnitudes of LFD impact on CO₂ become lesser in higher quantiles against lower ones. The quantile regression estimates of the effect of financial development (FD), renewable energy (REN), total natural resources rents (TNR), and GDP on carbon emissions (CO₂) are displayed in Figure 1 for each of the considered quantiles (0.25, 0.50, 0.75, and 0.90). This makes it possible to evaluate how the impacts of these variables change at different rates of carbon emissions across the sample of 11 emerging economies. That is, across quantiles, financial development, renewable energy consumption, natural resource rents, and GDP differently impact CO₂ emissions. Financial development has a positive impact on emissions in lower quantiles (the 25th and 50th percentiles), indicating that in the low emission economies, growth of the financial services boosts industrial activity and energy consumption. Nonetheless, at higher quantiles (75th and 90th percentiles), it loses its statistical significance, which might be due to less dirty investments or more stringent environmental measures. Renewable energy use decreases CO₂ emissions at all levels, but most significantly at the upper levels, showing that its use is most potent in regions with high emissions. Likewise, the impact of natural resource rents on emissions is greatly negative and becomes more pronounced at higher quantiles, indicating that good governance of resource windfalls and revenues can foster better environmental performance. GDP has a positive but statistically insignificant effect on emissions at all quantiles, bluntly contradicting the EKC’s claim that emissions grow as a country’s economy grows and only subsides at higher income levels.

The findings suggest that economic growth should be complemented by proactive sustainability policies rather than assumed to be self-correcting, and directing financial resources towards environmentally sustainable projects is crucial in managing the growth emissions trade-off. Additionally, a shift towards renewable energy and efficient resource management can drive sustainable development while maintaining economic competitiveness. Finally, the study provides strong economic evidence that financial development, renewable energy, and resource management play crucial roles in shaping environmental quality. Policymakers must adopt a differentiated approach, ensuring financial sector reforms, targeted renewable energy investments, and effective use of resource rents to achieve sustainable economic growth.

The Durbin–Wu–Hausman test was conducted to test for endogeneity in the model, as shown in

Table 7. Durbin–Wu–Hausman (DWH) test and causality tests.

Test	Chi-Square		Probability
DWH Test	1.23		0.287
Direction	W-Bar	Z-Bar	Probability
$lgdp\to lco2$	1.7229	1.6953	0.0900
$lfd\to lco2$	2.4114	6.8113	0.0000
$lren\to lco2$	3.6926	6.3146	0.0000
$ltnr\to lco2$	2.8752	5.6321	0.0000

, and the non-significant of the chi-square establishes that the explanatory variables are not correlated with the stochastic error term, suggesting endogeneity.

The MMQR result provides the influence of each variable on carbon emissions as shown in

Table 8. Regression analysis.

Dep. LCO2			Quantiles
	Location	Scale	0.25	0.50	0.75	0.90
LFD	1.1356 *** (0.3692)	−0.2635 (0.3038)	1.3508 *** (0.2006)	1.1733 *** (0.3326)	0.8741 (0.6458)	0.5694 (0.9907)
LREN	−0.9450 *** (0.2627)	−0.3685 * (0.2162)	−0.6441 *** (0.1438)	−0.8923 *** (0.2358)	−1.3106 *** (0.4513)	−1.7366 ** (0.7096)
LTNR	−1.5189 *** (0.2351)	−0.8302 *** (0.1935)	−0.8410 *** (0.1347)	−1.4002 *** (0.2095)	−2.3428 *** (0.3678)	−3.3024 *** (0.6628)
GDP	−0.2710 (0.4438)	−0.1551 (0.3652)	−0.1444 (0.2406)	−0.2488 (0.3997)	−0.4249 (0.7779)	−0.6041 (1.1888)
Cons	−10.1896 *** (1.8901)	6.0761 *** (1.5553)	−15.1511 *** (1.0693)	−11.0588 *** (1.7070)	−4.1602 (3.1154)	−2.8630 (5.2701)

Source: Author’s own work (2024). Note: The values in parentheses are the standard errors. ***, **, and * imply significance at the 1%, 5%, and 10% levels, respectively.

, however, there is still a need to check the causal relationship among the variables. Therefore, we employed the [58] causality analysis to further robustly check the results. The [58] causalty results are shown in Table 8, where we observe the association in relation to financial development (captured with domestic credit to private sector (LFD)), natural resources rent, renewable energy, economic growth, and with carbon emissions. The results indicate that any strategy focusing on these factors would considerably enhance the quality of the environment. Concentrating on these aspects to cut carbon emissions and enhance the quality of the environment will assist the 11 economies in accomplishing their sustainable development objectives.

5. Conclusions and Recommendation

The 2015 Paris Accord obligates participating nations to cut carbon emissions and adjust to the impact of renewable energy, GDP, natural resources rent, and domestic credit to the private sector, although, the empirical progress towards the realization of the Paris Accord has been increasing at a decreasing rate of achievement, and environmental quality remains a global phenomenon. A number of studies have been examined and carried out to unravel various factors that could influence and mitigate carbon emissions in the hopes to resolve its global uprise. This study therefore fills the gap by re-examining the asymmetric effect of economic growth on environmental quality, in the existence of GDP per capita, financial development (captured with domestic credit to private sector (LFD)), natural resources rent, renewable energy, and CO₂ for the 11 countries of Egypt, Bangladesh, South Korea, Mexico, Iran, Nigeria, Vietnam, Indonesia, Turkey, Pakistan, and the Philippines. Furthermore, the method of moment quantile regression econometric technique is employed for the re-examining the nexus among the study’s variables of interest. The cointegration analysis demonstrates that GDP per capita, LFD, renewable energy, natural resources rent imports, and CO₂ emissions have a long-run nexus with each other. Additionally, the results of the MMQREG analysis show that renewable energy and total natural resources rent significantly reduce CO₂ in 11 sampled countries, while on the contrary, financial development increases CO₂ emission in these countries, as GDPC depicts a negative but not significant influence on CO₂ in the case of 11 economies studied.

The natural resources rent and renewable energy may have good environmental quality, as they facilitate the transition to a low-carbon economy via redirecting towards low-carbon and sustainable development. Furthermore, to stimulate private sector investment in low-carbon and sustainable technology and enterprises, domestic credit with a lower interest rate should be used to encourage local investors into environmental quality investments that, in the long run, reduce carbon emission, as finance cum investment into this sector assist financial experts in enhancing their knowledge of environmental risks and possibilities. Finally, to reduce CO₂ emissions, it is imperative that 11 countries invest the proceeds from natural resources rent and renewable energy into renewable energy sources, such as wind power and solar power cum technologies aiding efficient environmental quality, thereby eliminating fossil fuel subsidies, as well as create support for low-carbon technology, thereby making the energy market more competitive.

Finally, the study is constrained by a lack of data, unaccounted variables such as quality of governance and innovation, possible causality issues, aggregation bias, and other methodological constraints which explain why future work needs to focus on governance, examine the adoption of clean technologies, analyze nonlinear interactions, perform disaggregated studies, and broaden the scope to include other environmental factors to fully understand the sustainability transitions.