Corporate social governance ESG policies and energy efficiency in G20 countries: The impact of policy coherence on environmental sustainability

Abstract

The increasing urgency for sustainable development in the face of climate change necessitates a deeper understanding of the interactions between corporate governance, environmental policies, and energy efficiency. Despite extensive research on corporate social responsibility (CSR) and sustainability, there remains a critical gap in assessing how these factors collectively influence ecological well-being in G20 nations. This study fills this gap by analyzing the relationships among ecological footprint (EFP), energy efficiency (EE), environmental patents (ENP), policy coherence (PC), economic growth (EG), and corporate governance (ESG and CSR) from 2000 to 2023. Utilizing advanced econometric techniques, including MMQR, FGLS, and Granger causality analysis, the findings provide significant insights. The results reveal that energy efficiency and environmental patents contribute to reducing ecological degradation, confirming their role in sustainable environmental outcomes. However, policy coherence does not exhibit a statistically significant impact on EFP, suggesting gaps in regulatory effectiveness. Moreover, CSR is negatively correlated with ecological degradation, indicating its beneficial role in improving environmental sustainability. Meanwhile, ESG practices demonstrate a significant positive relationship with ecological degradation, suggesting that current ESG strategies may not effectively contribute to environmental sustainability and may require stricter enforcement to prevent greenwashing. Additionally, economic growth (EG) is found to significantly reduce ecological degradation, highlighting that economic expansion in G20 nations might be linked to improved energy efficiency, technological innovation, or more effective environmental policies. These findings underscore the need for stronger environmental regulations, targeted energy policies, and corporate accountability to ensure meaningful sustainability progress. Policymakers must integrate enforceable sustainability measures into governance frameworks and enhance policy coherence to achieve tangible environmental benefits. Strengthening cross-sectoral collaboration and global partnerships will be crucial for aligning G20 nations with Sustainable Development Goals (SDGs), particularly in climate action and responsible consumption. This study contributes to the literature by offering a critical reassessment of how corporate governance, economic growth, and environmental policies influence sustainability, providing actionable insights for policymakers and stakeholders.

Keywords

sustainable development policy coherence energy efficiency corporate social governance ESG

Introduction

The urgent challenge of achieving sustainable development has risen to the forefront of global discourse in the twenty-first century, especially among G20 nations, which are crucial in shaping both the global economy and environmental policy.¹ As these countries navigate the complexities of rapid industrialization and urbanization, they face significant dilemmas in reconciling economic growth with environmental sustainability. This study investigates the intricate relationships among corporate governance, measured by Environmental, Social, and Governance (ESG) criteria, corporate social responsibility (CSR), environmental patents (ENP), energy efficiency (EE), and policy coherence (PC) concerning ecological footprints (EFP) across the G20 from 2000 to 2023.

Comprising approximately 80% of the global GDP and a substantial share of greenhouse gas emissions, the G20 nations play a pivotal role in the international effort to combat climate change.² The rapid pace of industrialization and evolving consumption patterns within these countries have resulted in increased carbon footprints, prompting urgent policy discussions centered on sustainable practices.^3–5 In response to these challenges, governments are increasingly implementing policies aimed at fostering environmental stringency and coherence.⁶ However, the efficacy of these initiatives is often undermined by a lack of alignment with corporate practices, highlighting the necessity for a cohesive strategy that integrates ESG considerations into corporate decision-making.⁷

ESG factors have gained prominence as a critical component of corporate governance, reflecting a firm's commitment to environmental stewardship, social responsibility, and sound governance practices. In G20 countries, ESG practices have been increasingly adopted in response to global initiatives like the Paris Agreement and the UN Sustainable Development Goals (SDGs).⁸ Studies by⁹ and¹⁰ emphasize that firms with robust ESG frameworks tend to exhibit better environmental performance, ultimately contributing to national and global sustainability efforts.

CSR represents a firm's voluntary commitment to social, environmental, and economic well-being. In the G20 context, CSR practices vary significantly owing to disparate regulatory frameworks, economic structures, and societal expectations. According to¹¹ and,¹² CSR initiatives in developed G20 nations tend to align with strategic business goals, whereas developing economies often emphasize community engagement and social welfare.

Environmental patents serve as indicators of a nation's commitment to green innovation and technological advancement. G20 countries, particularly those with high levels of industrial activity like the United States, China, and Germany, have witnessed a surge in ENP filings in response to environmental challenges.^13,14 The promotion of green patents not only addresses environmental concerns but also stimulates economic growth by fostering technological innovation.¹⁵

EE is a crucial factor in mitigating carbon emissions and boosting energy security. The G20 countries have implemented various EE measures, ranging from technological upgrades in manufacturing to policy-driven initiatives aimed at reducing energy consumption in residential sectors.¹⁶ Research by¹⁷ indicates that improving EE is essential for achieving long-term environmental goals, particularly in energy-intensive industries.

Policy coherence refers to the systematic alignment of various policy domains to achieve overarching goals. In the context of G20 nations, PC is vital for harmonizing environmental policies with economic and social objectives.¹⁸ Studies by¹⁹ and⁶ highlight that inconsistent policy frameworks can undermine environmental initiatives, resulting in suboptimal outcomes.

As the G20 grapples with these multifaceted challenges, this study seeks to bridge the existing gap in the literature by elucidating how the interplay among these variables influences ecological sustainability. Utilizing advanced econometric methodologies, the research aims to uncover causal relationships and offer actionable insights for policymakers and stakeholders. Ultimately, this study aspires to enrich the global dialogue on sustainable development, proposing strategies that facilitate a more harmonious balance between economic growth and environmental integrity within G20 nations.

This study will address several key inquiries, including: How do environmental patents (ENP) influence ecological footprints in G20 nations? What is the relationship between ESG and CSR in promoting sustainability? How does EE affect ecological degradation across different G20 countries? What role does PC play in harmonizing corporate practices with environmental goals? Are there significant causal relationships between these variables, and how do these relationships vary across different quantiles? Lastly, how can the findings inform policy recommendations for achieving sustainable development within the G20 context?

The remainder of the paper is organized in this manner. In Section 2, the relevant literature on environmental economics is summarized. Section 3 covers the main instruments used to compare global warming and environmental deterioration. Section 3 presents the data, and introduces the econometric model, and Section 4 presents our key insights. Conclusion and recommendations are provided in Section 5.

Literature review

This part is split into two primary segments: the first covers the theoretical background, and the second evaluates the relevant empirical literature.

Theoretical background

The theoretical background of this study centers on the complex interactions among economic growth, environmental sustainability, and corporate social responsibility (CSR), with particular attention to environmental, social, and governance (ESG) factors. The Environmental Kuznets Curve (EKC) hypothesis serves as a foundational framework, positing that environmental degradation initially increases with economic growth but later declines as income reaches a certain threshold, leading to enhanced environmental quality.²⁰ This dynamic underscores the need for transitioning from polluting industries to sustainable practices, especially as economies mature.^21,22

In this context, CSR emerges as a crucial element influencing both corporate behavior and environmental outcomes. By prioritizing sustainable practices, businesses can align their operations with broader societal goals, thereby mitigating their ecological footprints.¹¹ CSR initiatives often integrate ESG considerations, which evaluate a business's performance in terms of social responsibility, environmental stewardship, and governance systems. These factors have gained traction among investors and consumers alike, emphasizing the importance of transparency and accountability in achieving sustainability goals.⁷

Furthermore, the role of environmental technology and innovation cannot be overlooked. Investments in R&D and the adoption of clean technologies are essential for enhancing ecological performance and reducing emissions.²³ The presence of environmental patents reflects a country's capacity for innovation, enabling the diffusion of sustainable technologies that can significantly reduce ecological footprints.²⁴

This study aims to explore these theoretical frameworks to understand how economic growth, CSR, ESG factors, and technological innovation interconnect, particularly in the context of G20 nations, where the imperative for sustainable development is increasingly critical.^25,26 Ultimately, integrating these dimensions can provide valuable insights into achieving environmental sustainability while fostering economic development.

Empirical evidence

The empirical study that has evaluated the influences of environmental patents, EE, economic expansion, CSR, and ESG on GHG emissions in the backdrop of emerging countries is compiled sequentially in the following subsection.

Association between environmental quality and environmental patents

Lawmakers and ecological academics throughout the the globe are becoming increasingly aware of environmental innovation, thus it is critical to comprehend its unique characteristics when creating regulations.^27,28 Environmental new technologies are widely debated in the literature today concerning their potential to avoid pollution and safeguard environmental health.

Prior studies have focused on how people perceive the effects of the environment and how technology might help reduce CO₂ emissions.²⁹ Various factors, such as population expansion and technological progress, can lead to the deterioration of the environment.^30–32

Research by³³ asserts that patents should be granted for novel pollution-reduction techniques. A study by³⁴ illustrated the impact of carbon emissions and other levels of greenhouse gases within a country on patents and inventions about technologies addressing global warming. The volume of GHG production in a country influences the advancement and patenting of technologies connected to climate change. Brands serve as a crucial instrument for the production of technology and innovative products, representing the most effective approach to enhance these items.³⁵ Similarly,³⁶ assessed the favorable and substantial impacts of adaption technology and sustainable energy on eco-friendly quality in G8 countries over the period from 1990 to 2020.

H1:

The development and implementation of environmental patents reduce environmental degradation in G20 nations.

The connection between environmental quality and EE

Research by^37,38 examined the potential harm that energy use could have caused to the natural world in the G7 nations between 1970 and 2015. They found that while renewable energy reduces CO₂ emissions, expansion and EE increase them. Similarly,⁵ previously discovered that EE increases the release of CO₂ for the G7 countries while EE, institutional quality, and foreign direct investment (FDI) shorten contamination. They did this using the generalized method of moments (GMM) approach on data covering the years 1980–2014. The research conducted by³⁹ also demonstrated that FDI significantly enhances renewable energy use over the long run in China for the period from 1990 to 2021. Policies that incentivize and promote FDIs in renewable energy projects are crucial for enhancing the utilization of green energy and advancing toward more sustainable and ecologically benign energy sources.

On the other hand,⁴⁰ discovered that EE helped lower Egypt's ecological footprint levels based on annual data from 1971 to 2014. Comparably,⁴¹ also noted that the Belt and Road Initiative countries’ long-term carbon footprints have decreased as a result of efficient energy use. Similarly,⁴² said that because developing nations’ ecological laws and regulations are ineffective at preventing filthy foreign direct investment inflows, energy utilization raises greenhouse gas emissions statistics in these nations.

H2:

EE has a mixed impact on environmental quality in G20 nations, with potentially increasing environmental degradation.

Relationship between environmental quality and PC

According to previous research, policy coherence is believed to lower carbon emissions by incentivizing businesses and consumers to select eco-friendly products and lifestyles.⁴³ inspected the same relationship and found that carbon footprints and stringent environmental regulations are negatively correlated. Additionally,⁴⁴ uses a variety of seventy-seven regions of the Russian Federation to demonstrate the importance of stringent environmental legislation in lowering carbon emissions. An intriguing study conducted by⁴⁵ examined seven emerging countries and discovered an imbalanced relationship between carbon emissions and stringent environmental rules. The research conducted by⁴⁶ employing second-generation panel data methodologies and panel quantile regression demonstrated that natural resources augment carbon dioxide emissions in G-7 nations from 1990 to 2020. The combination of natural resources and strict environmental regulations markedly decreases emissions across many quantiles. Their results indicated that implementing stringent environmental rules is essential for attaining sustainable development goals (SDGs) associated with economic growth, innovation, sustainable urbanization, responsible resource consumption, and climate action.

H3:

PC improves ecological quality in G20 countries.

Nexus between environmental quality and CSR

Assert that corporate social responsibility (CSR) allows companies to secure bank loans at reduced interest rates and extended loan durations.⁴⁷ CSR refers to an organization's ability to engage in socially responsible actions that promote the growth and development of its surrounding environment or community.⁴⁸

The relationship between CSR and ecological viability has become increasingly significant as businesses seek to balance profitability with environmental stewardship. CSR refers to a company's ethical commitment to sustainable practices that benefit society and the environment. Conversely, ecological viability entails maintaining ecosystem functions, biodiversity, and resource sustainability. Research indicates that CSR positively impacts environmental performance. For instance, a study by⁴⁹ argues that integrating CSR into corporate strategies enhances both competitive advantage and environmental sustainability. Additionally,⁵⁰ found a strong correlation between CSR engagement and reduced ecological footprints, suggesting that firms committed to responsible practices tend to perform better environmentally.

The triple bottom line (TBL) framework underscores this connection, emphasizing the importance of social, environmental, and economic outcomes in assessing corporate success.⁵¹ Firms that prioritize CSR are more likely to adopt environmentally friendly technologies, enhancing their ecological viability.^52,53 further suggest that CSR initiatives, especially those focused on environmental issues, improve corporate credibility and stakeholder trust, fostering an environment conducive to sustainable practices.

Despite this growing body of literature, gaps remain in understanding the contextual factors that affect CSR's effectiveness across various industries and regions. Future research should explore these dynamics to provide deeper insights into how CSR can enhance ecological viability.

H4:

There is a favorable correlation between Corporate Social Responsibility (CSR) and ecological viability, indicating that higher CSR engagement correlates with improved environmental performance.

Nexus between environmental quality and ESG

The current literature on corporate ESG performance predominantly emphasizes the financial implications of ESG initiatives.⁵⁴ aggregated data from more than 2000 studies examining the correlation between ESG variables and corporate financial performance. Approximately 90% of the studies indicated a positive correlation between ESG characteristics and a company's financial performance. This illustrates the significance of sustainability issues for a company's enduring success. According to,⁵⁵ strong ESG performance can function as a safeguard, mitigate legal issues, and enhance profitability, consequently reducing bond spreads and corporate financial risk. They observed that firms with elevated ESG ratings are more inclined to reveal high-quality information, therefore reducing information asymmetry and bond spreads.⁵⁶ found that the effectiveness of labor investment in China might be augmented by higher ESG performance. Favorable ESG performance alleviates both excessive and insufficient investment in labor. Organizations demonstrating strong ESG performance can reduce human capital waste and alleviate labor shortages.

H5:

ESG would collaborate on environmental quality in the G20 nations.

Literature gaps and contributions

This study's originality and significance stem from its thorough examination of the interconnections among ESG standards, CSR, environmental patents, EE, policy consistency, and ecological footprints in G20 nations from 2000 to 2023. This study addresses a critical deficiency in the current literature by providing a comprehensive examination of the interplay between corporate governance, environmental innovation, and policy coherence in their aggregate impact on ecological sustainability in leading global economies.

This study employs a holistic approach, contrasting with previous research that typically examines these issues in isolation, so offering a more nuanced comprehension of the interplay between business and policy-driven sustainability programs. The study examines the significant disparity between business behavior and national environmental policies, highlighting the necessity for cohesive solutions to attain sustainable ecological equilibrium. This gives essential insights into how corporate governance and environmental innovation might promote sustainable development.

This study's primary innovation is in its utilization of sophisticated econometric tools, including MMQR, AMG, CCEMG, and FE models, to analyze both cross-sectional and temporal fluctuations in the data. The application of MMQR specifically facilitates the analysis of variable effects across several quantiles, uncovering heterogeneous consequences sometimes neglected by conventional mean-based models. This methodological improvement strengthens the findings and provides a more thorough knowledge of the links being examined.

Moreover, the study's emphasis on G20 nations enhances its originality, as these countries jointly account for a significant share of global GDP and carbon emissions, rendering the findings universally pertinent. This paper presents empirical evidence regarding the efficacy of policy coherence and corporate accountability in promoting ecological sustainability, offering practical insights for politicians, corporate executives, and researchers seeking to encourage sustainable development.

This work addresses deficiencies in the current literature and establishes a basis for future research to investigate analogous processes in other economic situations, especially in emerging economies where environmental and governance policies are swiftly advancing.

Data, model formulation, and approach

Data

This study analyzes annual panel data from 2000 to 2023 to examine the short- and long-term relationships between ecological footprint (EFP), environmental patents (ENP), policy coherence (PC), EE, ESG policies, CSR, and economic growth (EG) in G20 countries, excluding Argentina and Saudi Arabia due to data limitations in table 1. The focus on G20 nations stems from their significant global economic influence and contribution to environmental degradation. The study emphasizes the impact of ESG, CSR, EE, EG, and ENP on CO2 emissions, with EFP serving as the dependent variable. Data sources include the World Development Indicators (WDI) and OECD.

The descriptive statistics in Table 2 indicate substantial variation across the variables. EFP has a mean of 4.328, with a maximum value of 10.927 and a minimum of 0.084, suggesting a broad distribution. EG shows significant fluctuations, ranging from −10.624 to 13.636, with a relatively high standard deviation of 3.366. ENP exhibits the highest variability, with values spanning from 0 to 83.084, implying diverse environmental innovation levels across observations. PC and EE maintain moderate dispersion, while ESG and CSR display values near zero in mean and median, indicating a balance between positive and negative observations. The skewness and J-B statistics confirm non-normality in most variables, justifying the need for robust econometric techniques.

Table 1.

Variables and the data source.

Acronym	Variables	Proxy & Measurement	Type of variable	Source of data
EFP	Ecological Footprint	Ecological footprint (global hectares per person)	Dependent	GFN
EG	Economic Growth	GDP per capita growth (annual %)	Control	WDI
EE	Energy Efficiency	Energy intensity level of primary energy (MJ/ $2017 PPP GDP)	Independent	WDI
PC	Policy Coherence	Environmental policy stringency	Independent	OECD
ENP	Environmental Patents	Environmental technology patents (per person inventions)	Independent	OECD
CSR	Corporate Social Responsibility	PCA of Listed domestic companies and Research and development expenditure (% of GDP)	Independent	WDI
ESG	Environmental, Social, and Governance policies	PCA of the indicators of social factors (life expectancy at birth, the ratio of female to male labor force participation rate, current health expenditure, and employment to population ratio), environmental factors (forest area, renewable energy consumption, annual freshwater withdrawals, and level of water stress), and governance factors (regulatory quality, rule of law, voice and accountability, political stability and absence of violence/terrorism, government effectiveness, and control of corruption)	Independent	WDI

Here, OECD: Organisation for Economic Co-operation and Development, GFN: Global Footprint Network, and WDI: World Development Indicators.

Table 2.

Results of descriptive statistics.

	EFP	EG	ENP	PC	EE	ESG	CSR
Mean	4.328	2.207	13.021	2.026	4.997	−2.30E-9	1.37E-9
Median	4.268	1.922	9.430	2.026	4.435	−0.091	−1.91E-7
Maximum	10.927	13.636	83.084	4.889	12.060	1.569	3.521
Minimum	0.084	−10.624	0	0.1667	2.040	−1.688	−1.803
Std. Dev.	2.433	3.366	17.992	1.095	2.036	1.000	1.000
J-B stats	9.520***	19.230***	—	52.780***	39.500***	—	20.360***
Skewness	0.003***	0.038***	0.000***	0.394***	0.000***	0.528***	0.000***
Observations	408	408	408	408	408	408	408
Correlation analysis
EFP	1.000
EG	−0.241***	1.000
ENP	0.388***	−0.241***	1.000
PC	0.205***	−0.275***	0.567***	1.000
EE	0.096*	0.327***	−0.258***	−0.405***	1.000
ESG	0.599***	−0.290***	0.618***	0.544***	−0.203***	1.000
CSR	0.214***	−0.090***	0.608***	0.572***	−0.055	0.573***	1.000

Note: The significance at 1%, 5%, and 10% is denoted by ***, **, and *, respectively.

The correlation analysis reveals several notable relationships. EFP negatively correlates with EG (−0.241), implying that higher economic growth may contribute to environmental degradation. ENP and PC are positively associated (0.567), suggesting that increased environmental innovation aligns with stronger policy coherence. Similarly, ESG and CSR share a strong correlation (0.573), indicating a complementary role in corporate sustainability strategies. The positive relationship between ENP and EFP (0.388) suggests that environmental innovations contribute to ecological sustainability. However, the negative correlation between EG and PC (−0.275) signals potential trade-offs between economic expansion and policy coherence. EE exhibits a weak correlation with EFP (0.096), indicating that efficiency improvements alone may not significantly drive environmental performance. These findings highlight the complex interplay between sustainability factors and reinforce the need for integrated policy approaches.

Model formulation

Ecological footprints serve as the environmental degradation indicator in this research. The model adopted for this research is as follows:

E F P = f (E G, E N P, P C, E E, E S G, C S R)

(1)

The model examines the intricate linkages among essential economic and environmental variables and their influence on environmental degradation in G20 countries from 2000 to 2023. This study utilizes known theoretical frameworks to elucidate the interaction of these elements and their impact on environmental outcomes. The EKC hypothesis asserts an inverted U-shaped correlation between economic expansion and ecological degradation. Initially, environmental degradation escalates with economic expansion owing to industrialization; but, when attaining a specific income threshold, environmental quality is enhanced as societies adopt cleaner technology and enforce stronger laws. This theory advocates for the incorporation of the model to evaluate its differential effects on ecological footprints at various growth phases.⁵⁷ Moreover, the notion of innovation and technology dissemination supports the incorporation of environmental patents. Technological advancements, especially green technology, are essential for alleviating environmental degradation by enhancing manufacturing efficiency and diminishing emissions. The existence of patents indicates a nation's dedication to environmental sustainability and its capacity for innovation.⁵⁸ Institutional theory emphasizes the influence of regulatory frameworks on business and national conduct. Policy coherence is used to signify the efficacy of environmental policies in promoting sustainable practices. This hypothesis posits that nations with robust and cohesive environmental policies are more effective in diminishing ecological footprints.⁵⁹ EE is essential for mitigating environmental degradation through the optimization of energy consumption. This hypothesis posits that enhancements in energy intensity can substantially reduce carbon emissions while maintaining economic growth, consistent with the neoclassical growth model.⁶⁰ Freeman's stakeholder theory⁶¹ advocates for the incorporation of CSR and ESG factors. Companies using responsible practices consider the interests of diverse stakeholders, such as consumers, investors, and regulators, resulting in improved environmental outcomes.^62,63 CSR encompasses business efforts for societal benefit, whereas environmental, social, and governance (ESG) denotes a wider spectrum of practices related to environmental stewardship, social responsibility, and governance standards. The RBV theory supports the incorporation of knowledge assets, like as patents, by highlighting their strategic significance in achieving competitive advantage. Environmental innovation, as assessed by ENP, reflects a company's dedication to sustainable practices that can impact national environmental performance.⁶⁴ The theoretical discoveries result in the subsequent empirical model:

E F P_{i, t} = β_{0} + β_{1} E G_{i, t} + β_{2} E N P_{i, t} + β_{3} P C_{i, t} + β_{4} E E_{i, t} + β_{5} E S G_{i, t} + β_{6} C S R_{i, t} + ε_{i, t}

(2)

where EFP, EG, ENP, PC, EE, ESG, and CSR stand for the ecological footprint, economic growth, environmental patents, policy coherence, energy efficiency, environmental, social, and governance factors, and corporate social, responsibility factors, respectively.

The study strikes a compromise between complexity and clarity by retaining all variables in their original form. The degree of flexibility numbers β₁, β₂, β₃, β₄, β₅, and β₆ show the strength and direction of the relationship, whereas β₀ indicates the magnitude of the intercept. $ε_{i, t}$ delimits the error term in the regression model. The time series operator, which spans 2000 to 2023, is represented by the letter t. The cross-section, which is 17 G20 (except SAU and ARG), is represented by the letter i in the equation above.

Methodological approach

Principle component analysis (PCA)

The principal component analysis (PCA) is employed to create a composite index by combining multiple indicators into a single measure.⁶⁵ Here is a step-by-step explanation of the PCA process for the construction of CSR and ESG:

Standardizing the Variables: All variables are normalized to possess a mean of zero and a standard deviation of one to ensure comparability. The formula for standardization is:

Z = \frac{X - μ}{σ}

(3)

where X is the original value, μ is the mean, and σ is the standard deviation.

Computing the Covariance Matrix: A covariance matrix is computed to analyze the interrelationships among the variables. The covariance between two variables X and Y is computed as:

c o v (X, Y) = \frac{1}{n - 1} \sum_{i = 1}^{n} (X_{i} - \bar{X}) (Y_{i} - \bar{Y})

(4)

Eigenvalue and Eigenvector Calculation: Eigenvalues and eigenvectors are obtained from the covariance matrix. Eigenvalues represent the extent of variation elucidated by each principal component, whilst eigenvectors delineate the orientation of these components.

The equation for eigenvalues is:

A v = λ v

(5)

where A is the covariance matrix, λ is the eigenvalue, and ν is the eigenvector.

Selecting Principal Components: Components with eigenvalues greater than 1 are retained.

Computing the Principal Component Scores: Scores are determined by multiplying the standardized variables by their respective eigenvectors. The score for the ith component is defined as follows:

P C_{i} = \sum_{j = 1}^{p} Z_{j} \times v_{j, i}

(6)

Index Construction: The composite index is formulated as a weighted sum of the chosen primary components, with weights reflecting the variance attributed to each component:

I n d e x = \sum_{i = 1}^{k} λ_{i} \times P C_{i}

(7)

This PCA method guarantees a strong, data-driven strategy for index creation, yielding credible insights into the environmental performance of G20 nations.

Slope homogeneity test

The slope homogeneity test was used to determine whether the slope coefficients of the cointegration equation are homogenous. The test was first developed by,⁶⁶ but⁶⁷ expanded it and used it to obtain two statistics:

{\hat{Δ}}_{S - H T} = \sqrt{N} \times \sqrt{2 k} \times (\frac{1}{N} \hat{S} - k)

(8)

{\hat{\hat{Δ}}}_{a d j . S - H T} = \sqrt{N} \times \sqrt{\frac{T + 1}{2 k (T - k - 1)}} \times (\frac{1}{N} \hat{S} - 2 k)

(9)

where Ŝ and k represent Swamy's statistics and the independent variables, respectively, and T and N stand for time and country. For the test's p-value to validate the null hypothesis, it must be higher than 5%. The results of the homogeneity test developed by⁶⁷ are shown in Table 3.

Table 3.

Slope homogeneity test

Test	Value	P-value
${\hat{Δ}}_{S - H T}$	8.886\ast \ast \ast	0.000
${\hat{\hat{Δ}}}_{a d j . S - H T}$	10.883\ast \ast \ast	0.000

Note: The significance is shown as ***, **, and * at 1%, 5%, and 10%, respectively.

Cross-sectional dependence test

The CD test proposed by⁶⁸ can be used to approximate the cross-sectional measure of reliance on remains. This test enables us to determine the most appropriate panel unit root tests for examining the stationarity measures of the variables. The second-generation panel unit root analysis is appropriate if the calculated values of the residuals’ CD statistic are statistically more meaningful.

The CD assessment's equational form is presented below.

δ_{C - D P} = \frac{{((T \times N) (N - 1))}^{\frac{1}{2}}}{2} {\hat{P R}}_{N}

(10)

Second-generation unit root test

The G20 panel data were taken into account in this investigation. Panel data may also seem nonstationary, much like time-series data does. To avoid making inaccuracies in regressions, unit root analysis is crucial. Consequently, the CIPS and CADF tests from⁶⁸ are used in this investigation. Given is the CADF test.

Δ C S_{i, t} = φ_{i} + φ_{i} C S_{i, t - 1} + ϱ {\bar{C S}}_{t - 1} + \sum_{l = 0}^{p} ψ_{i, l} Δ {\bar{C S}}_{t - l} + \sum_{l = 1}^{p} υ_{i, l} Δ {\bar{C S}}_{i, t - l} + μ_{i, t}

(11)

Meanwhile, the CIPS is displayed as follows in Equation (5):

{\hat{C I P S}}_{U R} = \frac{1}{N} \sum_{l = 1}^{n} Ø_{i} (N, T)

(12)

According to these tests, the unit root test is repeated after calculating the variable's initial difference if one or more variables are not stationary. Precise testimony of cross-sectional dependency and heterogeneity is produced by these tests.

Quantile regression method

Developed the quantile-based technique to concentrate on the fixed effect in the scientific procedure, taking into account the many benefits of the MMQR.^69,70 The output generated by the quantile-based linkages between predictors and the dependent variables is represented by Equation (8) as the quantiles exposed to $Q_{y} (τ | X_{i, t})$ .

Y_{i, t} = α_{i} + X_{i, t}^{'} ψ + (δ_{i} + {Z^{'}}_{i, t} ϑ) U_{i, t}

(13)

The vectors shown in Equation (13) are thought to be known and further developed by Z, which stands for the transformation that is thought to happen as a result of differentiation and likewise has the modules l. It is important to keep in mind that $X_{i, t}$ always follows the same distribution for any cross-section (i) at any given time interval (t). Furthermore, every distinct cross-section (i) in a given time (t) can be represented by the distribution that $U_{i, t}$ employs. Thus, Equation (13) is rearranged even more to become Equation (15).

Z_{l} = Z_{l} (X), l = 1, 2, \dots ., k

(14)

Q_{y} (τ | X_{i, t}) = (α_{i} + δ_{i} (τ)) + X_{i, t}^{'} ψ + Z_{i, t}^{'} ϑ q (τ)

(15)

The importance of employing the MMQR method is in its capacity to encapsulate heterogeneity across various quantiles of the conditional distribution of the dependent variable. This is especially critical within the framework of the G20 nations, where environmental and economic dynamics can differ markedly among various countries and phases of development. Utilizing MMQR, the study may discern the differential effects of explanatory variables on ecological footprint across several quantiles, so offering a more refined comprehension of these associations. The study's brief duration, spanning from 2000 to 2023, renders MMQR especially pertinent as it can reveal distributional heterogeneities and short-term fluctuations that conventional mean-based approaches may neglect. A study by⁷¹ also examined the influence of environmental taxes, green finance, natural resources, human capital, and economic growth on environmental pollution in G20 nations from 2000 to 2022, using MMQR and other methodologies.

Robustness test

This study utilizes feasible generalised least squares (FGLS) as a robustness check to confirm the dependability of the findings. FGLS is especially adept at mitigating heteroskedasticity and autocorrelation problems prevalent in panel data analysis, rendering it a superior option for estimating models characterized by cross-sectional dependency. This method yields efficient and impartial estimates, even in the presence of heterogeneity or correlation in error terms across time and entities. In contrast to AMG and CCEMG, which emphasize unobserved common factors, FGLS addresses potential panel-level heteroskedasticity and adjusts standard errors accordingly, hence improving the accuracy of the computed coefficients. This work used FGLS to corroborate the major findings and assure robustness in assessing the correlations among EG, ENP, EE, CSR, ESG, PC, and EFP, due to its advantages over traditional fixed effects estimation.

The choice of these methodologies is also predicated on their capacity to address cross-sectional dependence and heterogeneity in panel data, which are essential factors due to the varied economic and environmental attributes of G20 nations. During the brief interval from 2000 to 2023, these methodologies effectively mitigate structural breaks, cross-sectional dependence, and potential endogeneity concerns, so assuring dependable inferences despite the constrained duration.

Granger causality test

One of the most well-known methods for determining whether or not there is a causative relationship between two variables in a panel data environment is the Granger-causality test, which was initially introduced by.⁷² Granger causality evaluations in panel data are useful for determining the trajectory of causality between variables across cross-sections (such as states or territories) with time. Granger causality is a method that investigates whether the past values of one variable may predict the current value of another variable while controlling for the previous values of the variable that is anticipated. This is in contrast to simple correlation, which merely displays the degree and direction of a link. When dealing with panel data, this is of utmost significance because it takes into consideration the heterogeneity that exists between units and dynamics over time.

By applying this test, the study can identify whether changes in ESG practices, EE improvements, or environmental patenting activities precede shifts in ecological footprints, thereby informing the development of more targeted and effective policy interventions. For short-term data, this test is crucial as it captures dynamic interactions and immediate causal effects that provide timely insights for policymakers.

Results and discussions

The results of the slope heterogeneity test in Table 3 show that both the ${\hat{Δ}}_{S - H T}$ statistic (8.886) and the ${\hat{\hat{Δ}}}_{a d j . S - H T}$ statistic (10.883) is highly significant at the 1% level, with p-values of 0.000. This indicates strong evidence of slope heterogeneity across the G20 countries in the study. In other words, the relationships between the variables, such as ecological footprint, environmental patents, policy coherence, EE, ESG policies, CSR, and economic growth, vary significantly across these countries, justifying the need for models that account for this heterogeneity. The present analysis utilizes the S-HT assays from.⁶⁷ The null hypothesis of slope homogeneity for both models is debunked by Table 3’s delta tilde and modified delta, which have strong significant probability values according to the S-HT test results. The paper discusses the observed empirical results in the following section.

The outcomes of the CD test in Table 4 indicate significant cross-sectional dependence among most variables, confirming that economic, environmental, and corporate sustainability factors are interconnected across the 17 G20 countries included in this study. The high CD values and p-values of 0.000 for EFP, EG, ENP, PC, EE, and CSR suggest strong interdependencies across G20 nations, implying that sustainability, corporate governance, and environmental policies are interconnected globally. Interestingly, ESG shows a negative CD value (−2.110) with a p-value of 0.034, indicating weaker but still statistically significant cross-sectional dependence. These findings justify the use of robust panel estimation techniques like FGLS, which account for cross-sectional dependence to ensure more reliable results.

Table 4.

Cross-section dependence (CD).

Variables	Value	P-value
EFP_i,t	6.770***	0.000
EG_i,t	33.630***	0.000
ENP_i,t	11.780***	0.000
PC_i,t	28.380***	0.000
EE_i,t	30.830***	0.000
ESG_i,t	−2.110**	0.034
CSR_i,t	3.730***	0.000

Note: The significance at 1% and 5% is specified by *** and **, respectively.

The unit root test results using CADF and CIPS methods reveal mixed stationarity properties across variables, indicating the necessity for first-difference transformation in most cases.

In table 5, For the CADF test, variables EG (−2.751) and ESG (−4.078) are stationary at level (I(0)) at the 5% and 1% significance levels, respectively, implying their mean-reverting nature without differencing. However, EFP (−1.742), ENP (−1.508), PC (−2.582), EE (−1.227), and CSR (−1.264) are non-stationary at level (I(0)), necessitating first-differencing for stationarity. After first differencing, EG (−4.225), PC (−3.728), and ESG (−4.568) become highly significant at 1%, confirming their stationarity at I(1). However, EFP (−1.957), ENP (−1.815), EE (−1.934), and CSR (−2.160) remain non-stationary even after first-differencing, indicating possible structural breaks or cross-sectional dependencies influencing these variables.

For the CIPS test, EG (−3.709), ENP (−3.257), and ESG (−4.175) are stationary at level (I(0)) at 1% significance, showing no need for differencing. In contrast, EFP (−2.064), PC (−2.281), EE (−1.306), and CSR (−1.138) are non-stationary at level but achieve stationarity at I(1) with significant values after differencing. The higher absolute t-values at the first difference (e.g., EG −5.615, ESG −6.146) confirm strong stationarity at I(1). This behavior corresponds with the literature, including,⁶⁸ which emphasizes the necessity of first differencing in macroeconomic datasets to attain stationarity.

Overall, these results indicate that EG and ESG are consistently stationary at I(0) across both tests, while EFP, PC, EE, and CSR require first-differencing for stationarity under CIPS but not under CADF. The persistence of non-stationarity in EFP, ENP, EE, and CSR under CADF suggests potential heterogeneity or structural factors affecting their stochastic properties, which must be addressed through appropriate panel estimation techniques like FGLS to ensure reliable inferences.

In table 6, Westerlund's⁷³ cointegration test examines whether a long-run equilibrium relationship exists among the study variables. The test statistics Gt (−1.649), Ga (−16.297), Pt (−5.133), and Pa (−6.382) indicate weak evidence of cointegration. However, the high Z-values (e.g., Gt: 5.092, Pt: 5.100) and insignificant P-values (all > 0.05, with some reaching 1.000) strongly suggest that the null hypothesis of no cointegration cannot be rejected.

This means that despite potential long-run relationships, the data does not provide sufficient statistical evidence to confirm strong panel cointegration among the variables. The high P-values (e.g., Gt: 1.000, Pt: 1.000) imply that short-term fluctuations may dominate over long-term equilibrium trends, potentially due to structural breaks, cross-sectional dependencies, or heterogeneity in the panel. These findings also suggest that alternative estimation methods should be employed to address potential heteroskedasticity and cross-sectional dependence, ensuring robust long-run inference despite weak cointegration signals.

The MMQR estimates in Table 7 provide critical insights into the relationship between economic growth (EG) and ecological footprint (EFP) across different quantiles. The negative and statistically significant coefficient of EG in the location estimates (−0.093 at 1%) suggests that economic growth reduces EFP on average, supporting the Environmental Kuznets Curve (EKC) hypothesis,⁵⁷ which argues that economic expansion initially worsens environmental quality but later leads to improvements through better policies and technological advancements. Additionally, the scale parameter (−0.038 at 5%) indicates that EG also affects the variability of EFP across quantiles (see Figure 1).

Figure 1.

Graphical representation of MMQR.

Table 5.

Unit root tests.

Variables	Level (I(0))	1^st Difference (I(1))
Cross-Sectionally Augmented Dickey-Fuller (CADF)
EFP_i,t	−1.742	−1.957
EG_i,t	−2.751**	−4.225***
ENP_i,t	−1.508	−1.815
PC_i,t	−2.582	−3.728***
EE_i,t	−1.227	−1.934
ESG_i,t	−4.078***	−4.568***
CSR_i,t	−1.264	−2.160
Cross-Sectionally Augmented IPS (CIPS)
EFP_i,t	−2.064	−2.894***
EG_i,t	−3.709***	−5.615***
ENP_i,t	−3.257***	−3.892***
PC_i,t	−2.281**	−4.345***
EE_i,t	−1.306	−2.931***
ESG_i,t	−4.175***	−6.146***
CSR_i,t	−1.138	−3.646***

Note: *** and ** show the significance at 1% and 5%, respectively.

Table 6.

Cointegration test.

Statistics	Gt	Ga	Pt	Pa
Value	−1.649	−16.297	−5.133	−6.382
Z-Value	5.092	0.357	5.100	3.177
P-Value	1.000	0.640	1.000	0.999

Table 7.

MMQR estimates.

Variables	Coefficient	Standard error	Z-value	P-value
Location
EG_i,t	−0.093***	0.031	−3.01	0.003
ENP_i,t	0.028***	0.008	3.51	0.000
PC_i,t	−0.083	0.112	−0.74	0.457
EE_i,t	0.355***	0.060	5.96	0.000
ESG_i,t	1.627***	0.131	12.47	0.000
CSR_i,t	−0.652***	0.170	−3.85	0.000
Constant	2.566***	0.443	5.80	0.000
Scale
EG_i,t	−0.038**	0.018	−2.07	0.038
ENP_i,t	−0.003	0.005	−0.70	0.486
PC_i,t	−0.060	0.066	−1.91	0.365
EE_i,t	0.201***	0.035	5.70	0.000
ESG_i,t	−0.079	0.077	−1.02	0.309
CSR_i,t	0.304***	0.100	3.03	0.002
Constant	0.629**	0.262	2.40	0.016
Q_0.25
EG_i,t	−0.061*	0.032	−1.93	0.054
ENP_i,t	0.031***	0.008	3.78	0.000
PC_i,t	−0.032	0.114	−0.28	0.778
EE_i,t	0.185***	0.062	2.97	0.003
ESG_i,t	1.694***	0.133	12.70	0.000
CSR_i,t	−0.911***	0.174	−5.23	0.000
Constant	2.032***	0.453	4.49	0.000
Q_0.50
EG_i,t	−0.091***	0.031	−2.94	0.003
ENP_i,t	0.028***	0.008	3.57	0.000
PC_i,t	−0.079	0.111	−0.71	0.477
EE_i,t	0.341***	0.061	5.62	0.000
ESG_i,t	1.632***	0.130	12.60	0.000
CSR_i,t	−0.674***	0.169	−3.98	0.000
Constant	2.522***	0.440	5.73	0.000
Q_0.75
EG_i,t	−0.126***	0.038	−3.34	0.001
ENP_i,t	0.025***	0.010	2.60	0.009
PC_i,t	−0.135	0.136	−0.99	0.321
EE_i,t	0.528***	0.075	7.01	0.000
ESG_i,t	1.559***	0.159	9.83	0.000
CSR_i,t	−0.391*	0.208	−1.88	0.061
Constant	3.108***	0.538	5.78	0.000
Q_0.90
EG_i,t	−0.162***	0.050	−3.25	0.001
ENP_i,t	0.022*	0.013	1.72	0.085
PC_i,t	−0.192	0.180	−1.06	0.287
EE_i,t	0.718***	0.095	7.59	0.000
ESG_i,t	1.485***	0.210	7.09	0.000
CSR_i,t	−0.104	0.271	−0.38	0.702
Constant	3.701***	0.713	5.19	0.000

Note: The significance threshold is shown as * < 10%, ** < 5%, and *** < 1%. The dependent variable is ecological footprint (EFP).

At the lower quantiles (Q0.25 and Q0.50), EG maintains a weaker but still negative impact (−0.061 at 10% and −0.091 at 1%), implying that in economies with lower ecological footprints, economic growth has a mild role in reducing environmental pressures. However, at higher quantiles (Q0.75 and Q0.90), the negative effect strengthens (−0.126 and −0.162, both at 1%), suggesting that in high ecological footprint economies, growth significantly contributes to environmental relief, possibly due to sustainable technological transitions.

Regarding policy coherence (PC), its impact remains statistically insignificant across all quantiles, indicating that government policy interventions may not have a direct influence on EFP. The location coefficient (−0.083, p = 0.457) and scale parameter (−0.060, p = 0.365) suggest that PC does not significantly alter ecological footprint outcomes. Similarly, its quantile estimates remain non-significant, reinforcing that existing policy measures may lack the effectiveness needed to drive substantial environmental changes.

Other variables exhibit meaningful effects on EFP. ENP (0.028), EE (0.355), and ESG (1.627) positively impact EFP, highlighting that environmental patents, energy use, and environmental governance shape ecological trends. This finding of ENP opposes,³⁶ who underscored the importance of green technologies in addressing environmental issues. EE's finding aligns with,⁷⁴ who examined the rebound effects linked to EE initiatives. Conversely, CSR (−0.652) significantly reduces EFP, underscoring the role of corporate responsibility in environmental sustainability. These findings confirm that economic and environmental dynamics are highly quantile-dependent, requiring targeted policies for different ecological footprint levels. This discovery corresponds with,⁷ who emphasized the significance of corporate governance policies in promoting environmental sustainability. CSR indicators demonstrate a minimal adverse effect across all quantiles, suggesting that corporate social responsibility initiatives may necessitate more focused execution to achieve substantial ecological advantages, as indicated by.¹¹

The MMQR results in Figure 1 illustrate the heterogeneous impacts of key variables on ecological degradation across different quantiles. Economic growth (EG) exhibits a declining trend across quantiles, indicating that its mitigating effect on ecological degradation is more pronounced at higher pollution levels. Environmental patents (ENP) show a strong negative impact across all quantiles, with a steeper decline at upper quantiles, reinforcing their role in environmental sustainability. Policy coherence (PC) follows an inverted U-shape, initially increasing degradation at lower quantiles but reducing it at higher quantiles, suggesting its effectiveness strengthens in severely degraded environments. EE consistently shows a positive impact, with its effect amplifying at higher quantiles, highlighting its crucial role in pollution reduction. ESG practices (ESG) exhibit a positive association with ecological degradation, indicating potential inefficiencies or weak enforcement in environmental governance, especially at upper quantiles. Lastly, corporate social responsibility (CSR) demonstrates a rising negative impact at higher quantiles, implying that CSR initiatives become more effective in heavily polluted contexts. These findings emphasize the necessity for quantile-specific policies tailored to varying levels of ecological degradation.

The Feasible Generalized Least Squares (FGLS) results in Table 8 confirm the MMQR estimates, reinforcing the robustness of the findings. The negative and significant coefficient of EG (−0.094 at 1%) further supports the Environmental Kuznets Curve (EKC) hypothesis, indicating that economic growth contributes to reducing the ecological footprint (EFP), likely through sustainable policies and technological advancements in higher-growth economies. Similarly, ENP (0.028 at 1%) positively impacts EFP, aligning with previous results that suggest energy production increases environmental stress.

Table 8.

FGLS robustness test

Variables	Coefficient	Standard error	Z-value	P-value
EG_i,t	−0.094***	0.029	−3.28	0.001
ENP_i,t	0.028***	0.007	3.97	0.000
PC_i,t	−0.083	0.115	−0.72	0.469
EE_i,t	0.355***	0.050	7.11	0.000
ESG_i,t	1.627***	0.122	13.36	0.000
CSR_i,t	−0.652***	0.127	−5.15	0.000
Constant	2.567***	0.417	6.16	0.000
Wald test	368.830***	—	—	0.000

Note: The significance threshold is shown as*** < 1%, dependent variable: Ecological footprint (EFP).

The insignificance of PC (−0.083, p = 0.469) remains consistent with MMQR, indicating that policy controls may not have a substantial direct impact on EFP. This aligns with the argument that policy interventions may be ineffective or insufficiently enforced to drive significant ecological improvements.

The positive effect of EE (0.355 at 1%) indicates that EE plays a crucial role in influencing environmental outcomes, supporting theories that energy-intensive economies face greater ecological burdens. Likewise, the strong significance of ESG (1.627 at 1%) confirms that environmental, social, and governance factors directly shape sustainability trends, mirroring the MMQR results. The negative impact of CSR (−0.652 at 1%) further validates corporate social responsibility's role in mitigating environmental degradation.

The highly significant Wald test (368.830, p = 0.000) suggests the overall model is strongly explanatory. Given that FGLS mitigates heteroskedasticity and autocorrelation issues, these findings solidify the robustness of the MMQR conclusions, emphasizing the consistency of the economic and environmental relationships identified in this study.

The study's empirical findings provide mixed support for the proposed hypotheses. H1, which posits that environmental patents (ENP) reduce environmental degradation, is not supported. The results indicate a positive relationship between ENP and EFP, suggesting that rather than mitigating environmental degradation, increased environmental patents may be associated with higher ecological footprints. This could be due to the lag effect of patent implementation, where innovations take time to yield tangible environmental benefits, or the possibility that patents are being used more for economic gains than for genuine sustainability improvements. H2, which suggests that EE has a mixed impact on environmental quality, is supported. The findings reveal that EE is positively associated with EFP across different quantiles, implying that while EE improvements are intended to reduce environmental harm, they may lead to rebound effects where increased efficiency results in greater energy consumption, ultimately exacerbating environmental degradation. This aligns with prior research suggesting that efficiency gains alone are not sufficient to ensure ecological improvements. H3, which hypothesizes that policy coherence (PC) improves environmental quality, is not supported. The results show that PC does not have a significant impact on environmental degradation, indicating that greater alignment between policies does not necessarily lead to ecological improvements. This contradicts the theoretical expectation that coherent policies would enhance sustainability by integrating environmental goals into economic and social policies. The lack of significance may suggest weak enforcement mechanisms, conflicting policy objectives, or the need for stronger institutional frameworks to ensure effective implementation.

H4, which states that CSR improves environmental performance, is not supported. The results reveal a negative relationship between CSR and EFP, implying that higher corporate social responsibility engagement does not necessarily lead to better environmental outcomes. This may be due to ineffective CSR initiatives, greenwashing, or a focus on social and governance aspects over concrete environmental actions. H5, which argues that ESG promotes environmental quality, is not supported. The findings indicate a positive relationship between ESG and EFP, suggesting that rather than enhancing environmental sustainability, higher ESG commitments may be linked to increased environmental degradation. This could reflect weak enforcement of ESG policies, an overemphasis on governance and social aspects at the expense of environmental priorities, or ineffective environmental management practices within ESG frameworks. Overall, these findings highlight that while environmental policies and corporate strategies aim to improve ecological conditions, their real-world impact is often complex and may not always align with theoretical expectations. While EE and environmental patents hold potential for sustainability, their current implementation may not be effectively addressing environmental degradation in G20 nations.

The Granger causality analysis in Table 9 provides key insights into the directional relationships between EFP and its potential determinants. The results show that EE Granger causes EFP (F = 3.163, p = 0.043), suggesting that changes in EE significantly influence environmental degradation. However, the reverse causality is not observed (F = 0.506, p = 0.603), indicating that environmental degradation does not impact EE. Similarly, environmental patents (ENP) Granger-cause EFP (F = 4.513, p = 0.012), meaning that shifts in environmental patent activities contribute to environmental degradation, but EFP does not influence ENP (F = 0.904, p = 0.406).

Table 9.

Granger-causality analysis.

Null hypothesis	F-statistic	P-Value
EE_i,t→EFP_i,t	3.163**	0.043
EFP_i,t→EE_i,t	0.506	0.603
ENP_i,t→EFP_i,t	4.513**	0.012
EFP_i,t→ENP_i,t	0.904	0.406
CSR_i,t→EFP_i,t	1.913	0.149
EFP_i,t→CSR_i,t	0.128	0.880
EG_i,t→EFP_i,t	0.476	0.622
EFP_i,t→EG_i,t	8.818***	0.000
ESG_i,t→EFP_i,t	0.097	0.907
EFP_i,t→ESG_i,t	3.941**	0.020
PC_i,t→EFP_i,t	0.020	0.980
EFP_i,t→PC_i,t	1.490	0.227
CSR_i,t→ENP_i,t	0.184	0.832
ENP_i,t→CSR_i,t	16.557***	0.000

Interestingly, EFP Granger-causes economic growth (EG) (F = 8.818, p = 0.000), but EG does not Granger-cause EFP (F = 0.476, p = 0.622), implying that worsening environmental degradation may drive economic activities rather than the other way around. A unidirectional causality is also observed from EFP to ESG (F = 3.941, p = 0.020), meaning that environmental degradation influences ESG commitments, potentially indicating a reactive approach by businesses and policymakers rather than proactive environmental governance.

No significant causal relationships are found between CSR and EFP, suggesting that CSR activities do not directly influence environmental degradation (F = 1.913, p = 0.149) and vice versa (F = 0.128, p = 0.880). Similarly, policy coherence (PC) shows no significant causality with EFP in either direction. However, a strong causal link is found from ENP to CSR (F = 16.557, p = 0.000), suggesting that environmental patent activities significantly shape CSR engagement, possibly due to regulatory requirements or stakeholder expectations.

These findings indicate that technological and policy-related factors, such as EE and ENP, play crucial roles in driving environmental degradation, while economic and corporate governance aspects seem to respond to worsening environmental conditions rather than proactively influencing them.

Conclusion

This study investigates the interplay between environmental policies, EE, corporate social responsibility, and ecological footprints in G20 nations, excluding Argentina and Saudi Arabia, from 2000 to 2023. Employing various econometric techniques, including MMQR, FGLS, and Granger causality, the study aims to provide a comprehensive understanding of these relationships. The MMQR results indicate that EE exacerbates EFP, contradicting expectations, while environmental patents (ENP) significantly contribute to increased ecological footprints rather than reducing them. Policy coherence (PC) shows no significant impact, failing to support its hypothesized role in improving ecological sustainability. Corporate social responsibility (CSR) negatively affects environmental quality, suggesting that CSR efforts may be insufficient or ineffective in curbing environmental degradation.

Conversely, ESG factors (ESG) demonstrate a positive effect on EFP, indicating that increased ESG engagement does not necessarily lead to environmental improvements. The FGLS robustness tests further validate these findings, with EE consistently showing a significant positive effect on ecological footprints, reinforcing the notion that improvements in energy use do not necessarily translate into better environmental quality.

The Granger causality analysis highlights that EE and ENP drive ecological degradation rather than mitigate it. Additionally, EFP Granger causes ESG commitments, indicating that worsening environmental conditions prompt ESG engagement rather than ESG proactively reducing degradation. A unidirectional causality from EFP to economic growth (EG) suggests that environmental degradation may stimulate economic activities rather than the other way around. Overall, the study underscores the unexpected adverse effects of EE and environmental innovation on ecological degradation and the reactive nature of ESG and corporate responsibility efforts. These insights provide valuable considerations for policymakers and stakeholders in refining environmental policies and corporate sustainability strategies.

Policy implications

This study's findings offer critical policy implications for promoting sustainable development in G20 nations and beyond. To achieve meaningful environmental progress, policymakers must adopt a data-driven approach aligned with the United Nations Sustainable Development Goals (SDGs) while addressing the unintended consequences of certain sustainability strategies identified in this research.

First, EE policies need reassessment, as findings indicate that increased EE is linked to greater ecological degradation rather than its reduction. This suggests the potential for a rebound effect, where efficiency gains lower costs and leads to increased energy consumption. Policymakers should introduce demand-side management policies, carbon pricing mechanisms, and stricter efficiency standards that discourage excessive energy use. Additionally, integrating renewable energy sources with efficiency measures is crucial to ensuring actual environmental benefits, contributing to SDG 7 (Affordable and Clean Energy).

Second, environmental innovation through patents (ENP) does not currently yield the expected ecological benefits, as it appears to be associated with increasing environmental degradation rather than its reduction. Policymakers must evaluate the nature and implementation of environmental patents, prioritizing genuinely sustainable innovations over patents that promote efficiency without reducing emissions. Strengthening green patent approval criteria, increasing direct funding for breakthrough clean technologies, and ensuring technology transfer to industries with high environmental impact can better align with SDG 9 (Industry, Innovation, and Infrastructure).

Third, policy coherence (PC) was found to have no significant impact on ecological footprints, suggesting that current policy frameworks may lack effective enforcement or integration. To enhance environmental governance, inter-ministerial coordination, cross-sectoral environmental policies, and legally binding sustainability targets should be established. Governments should regularly assess policy effectiveness and eliminate regulatory contradictions that hinder environmental progress, supporting SDG 13 (Climate Action).

Fourth, corporate social responsibility (CSR) has a negative relationship with environmental quality, indicating that voluntary corporate actions alone are insufficient to drive sustainable outcomes. Policymakers should enforce stricter corporate environmental regulations, introduce mandatory sustainability reporting standards, and link CSR initiatives to measurable ecological impact. Establishing green tax incentives tied to verifiable reductions in ecological footprints can enhance CSR's effectiveness, aligning with SDG 12 (Responsible Consumption and Production).

Fifth, ESG (Environmental, Social, and Governance) factors, contrary to expectations, show a positive association with ecological degradation (EFP), suggesting that ESG commitments may be reactive rather than proactive in improving environmental quality. Policymakers should introduce standardized ESG regulations, eliminate greenwashing, and ensure that ESG investments prioritize tangible environmental improvements. Strengthening accountability mechanisms for ESG-driven corporate actions can enhance their contribution to sustainability, supporting SDG 17 (Partnerships for the Goals).

Lastly, international cooperation must be expanded beyond voluntary agreements. G20 nations should lead binding climate commitments, technology-sharing agreements, and financial support for transitioning economies to adopt sustainable technologies. Establishing global regulatory frameworks for corporate sustainability disclosures and carbon trading mechanisms can create a more effective international environmental strategy. Collaboration with global institutions like the UN and World Bank remains critical for advancing SDG 17.

To ensure these policies are effective, systematic monitoring, data-driven adjustments, and stringent enforcement mechanisms must be implemented. G20 nations must take a leadership role in setting higher environmental standards, creating a global precedent for sustainable and resilient development.

Limitations and future directions

The study acknowledges several limitations. Firstly, the focus is exclusively on G20 countries, excluding nations like Argentina and Saudi Arabia due to data unavailability, which may affect the generalizability of the findings. Secondly, while the analysis utilizes robust econometric techniques, the reliance on available data may introduce measurement errors or biases, particularly in self-reported metrics like corporate social responsibility (CSR). Additionally, the study covers the period from 2000 to 2023, potentially limiting insights into recent developments and emerging trends in sustainability practices.

Future research directions should consider expanding the geographical scope beyond G20 nations to include developing and smaller economies, offering a more holistic understanding of ecological policies’ impact on sustainability across diverse economic landscapes. Longitudinal studies can provide valuable insights into the dynamic relationships between EE, environmental patents, and corporate governance practices over time. Additionally, integrating qualitative approaches, such as case studies and expert interviews, can deepen the contextual understanding of how policy frameworks and corporate strategies shape environmental outcomes. From a methodological perspective, future research should explore additional robustness checks by employing techniques like Ordinary Least Squares (OLS), Generalized Method of Moments (GMM), and Panel Vector Autoregression (PVAR) to validate findings further and address potential endogeneity issues. Incorporating these techniques will complement the already applied AMG, CCEMG, and fixed-effects models, ensuring a more comprehensive and resilient analysis. Finally, investigating the synergies between technological innovations, climate change policies, and sustainability indicators can offer actionable insights for policymakers striving to achieve Sustainable Development Goals (SDGs) effectively.

Footnotes

Credit authorship contribution statement

PEI pei: Writing – review & editing, Writing – original draft, Methodology, Data curation.

Xinxue Chang : Writing – review & editing, Writing – original draft, Methodology, Data curation.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article This work was supported by National Natural Science Foundation of China (71872040) and the National Social Science Foundation of China under Grant [National Office for Philosophy and Social Sciences] (19ZDA097).

National Social Science Foundation of China under Grant [National Office for Philosophy and Social Sciences], National Natural Science Foundation of China, (grant number 19ZDA097, 71872040).

ORCID iD

Pei Pei

Pei Pei is a PhD candidate at the School of Economics and Management, Southeast University. He is also a vice professor in Lanzhou University of Technology. His research focuses on social responsibility and corporate value creation, management decision-making and system analysis.

Xinxue Chang is a PhD candidate at the School of Economics and Management, Southeast University. Her research focuses on corporate social responsibility and financial management.

References

OECD. G20 Contribution to the 2030 Agenda. G20 Contrib to 2030 Agenda [Internet]. 2019 Sep 7 [cited 2025 Jun 3]; Available from: https://www.oecd.org/en/publications/g20-contribution-to-the-2030-agenda_db84dfca-en.html

Ali

Shayanmehr

Radmehr

, et al. Investigating environmental quality among G20 nations: the impacts of environmental goods and low-carbon technologies in mitigating environmental degradation. Geosci Front [Internet]. 2024 [cited 2025 Jun 3];15:101695. Available from: https://www-sciencedirect-com-443.web.bisu.edu.cn/science/article/pii/S1674987123001627

Usman

Balsalobre-Lorente

. Environmental concern in the era of industrialization: can financial development, renewable energy and natural resources alleviate some load? Energy Policy 2022; 162: 112780.

Figueres

Le Quéré

Mahindra

, et al. Emissions are still rising: ramp up the cuts. Nature 2018; 564: 27–30.

Shahbaz

Balsalobre

Shahzad

SJH

. The influencing factors of CO2 emissions and the role of biomass energy consumption: statistical experience from G-7 countries. Environ Model Assess 2019; 24: 143–161.

van der Voet

. Policy preferences in response to negative economic prospects of COVID-19: a survey-experiment among local politicians in four European countries. Public Adm 2022; 100: 356–374.

Eccles

Ioannou

Serafeim

. The impact of corporate sustainability on organizational processes and performance. Manage Sci 2014; 60: 2835–2857.

UNEP. Sustainable Finance Policy: the role of G7 and G20 | Policy report | PRI [Internet]. 2021 [cited 2025 Feb 17]. Available from: https://www.unpri.org/policy/sustainable-finance-policy-the-role-of-g7-and-g20/8646.article

Brogi

Lagasio

Porretta

. Be good to be wise: environmental, social, and governance awareness as a potential credit risk mitigation factor. J Int Financ Manag Account 2022; 33: 522–547.

10.

Devalle

Fiandrino

ESG Sustainability and financial capital structure: where they stand nowadays valter cantino full professor of accounting. Int J Bus Soc Sci [Internet]. 2017 [cited 2025 Feb17];8:23–26. Available from: https://www.unpri.org/policy/sustainable-finance-policy-the-role-of-g7-and-g20/8646.article

11.

Carroll

Shabana

. The business case for corporate social responsibility: a review of concepts, research and practice. Int J Manag Rev 2010; 12: 85–105.

12.

Matten

Moon

Reflections on the 2018 decade award: the meaning and dynamics of corporate social responsibility. Acad Manag Rev [Internet]. 2020 [cited 2025 Feb 17];45:7–28. Available from: https://research.cbs.dk/en/publications/reflections-on-the-2018-decade-award-the-meaning-and-dynamics-of-,

13.

Popp

Environmental policy and innovation: a decade of research. 2019 [cited 2025 Feb 17]; Available from: https://www.nber.org/papers/w25631

14.

Pan

Jiang

Wang

. Does environmental regulation promote green technological innovation of companies? Evidence from green patents of Chinese listed companies. International J Low-Carbon Technol [Internet]. 2024 [cited 2025 Feb 17];20:543–552. Available from: https://academic.oup.com/ijlct/article-abstract/doi/10.1093/ijlct/ctad078/7638544

15.

Ahmed

Ahmad

Rjoub

, et al.

Economic growth, renewable energy consumption, and ecological footprint: exploring the role of environmental regulations and democracy in sustainable

. Sustainable Dev [Internet]. 2021 [cited 2025 Feb 17];30:595–605. Available from: https://onlinelibrary-wiley-com-443.web.bisu.edu.cn/doi/abs/10.1002/sd.2251

16.

IEA. Energy Efficiency 2020 – Analysis - IEA [Internet]. 2020 [cited 2025 Feb 17]. Available from: https://www.iea.org/reports/energy-efficiency-2020

17.

Huang

Miao

Wei

. Rebound effects of energy efficiency improvement based on computable general equilibrium models: a systematic review. Energy Effic 2023: 16: 662–669.

18.

OECD. Driving Policy Coherence for Sustainable Development. Driv Policy Coherence Sustain Dev. 2023.

19.

Waheed

Fischer

Khan

Climate change policy coherence across policies, plans, and strategies in Pakistan—implications for the China–Pakistan economic corridor plan. Environmental Manag [Internet]. 2021 [cited 2025 Feb 17];67:793–810. Available from: https://link-springer-com-443.web.bisu.edu.cn/article/10.1007/s00267-021-01449-y

20.

Grossman

Krueger

Environmental impacts of a North American free trade agreement. 1991 Nov [cited 2024 May 6]; Available from: http://www.nber.org/papers/w3914.pdf

21.

Balsalobre-Lorente

Shahbaz

Chiappetta Jabbour

, et al. The role of energy innovation and corruption in carbon emissions: evidence based on the EKC hypothesis. Green Energy Technol 2019; 32: 271–304.

22.

Tenaw

Beyene

Environmental sustainability and economic development in sub-Saharan Africa: A modified EKC hypothesis. Renewable Sustain Energy Rev [Internet]. 2021 [cited 2024 May 6];46:543–550. Available from: https://www-sciencedirect-com-443.web.bisu.edu.cn/science/article/pii/S136403212100191X doi:https://doi.org/10.1016/j.rser.2021.110897

23.

Davidson

Exnovating for a renewable energy transition. Nature Energy [Internet]. 2019 [cited 2024 May 6];67:1001–1012. Available from: https://www-nature-com-s.web.bisu.edu.cn/articles/s41560-019-0369-3

24.

Ahmed

Cary

Ali

, et al. Moving toward a green revolution in Japan: symmetric and asymmetric relationships among clean energy technology development investments, economic growth, and CO2 emissions. Energy Environ 2021; 33: 1417–1440.

25.

Ahmad

Jabeen

Heterogeneity of pollution haven/halo hypothesis and environmental Kuznets curve hypothesis across development levels of Chinese provinces. Journal Clean Prod [Internet]. 2021 [cited 2024 May 6];73:348–356. Available from: https://www-sciencedirect-com-443.web.bisu.edu.cn/science/article/pii/S0959652620349428

26.

Wang

Sun

, et al.

Resource abundance, industrial structure, and regional carbon emissions efficiency in China.

Resources Policy [Internet]. 2019 [cited 2024 May 6];54:673–680. Available from: https://www-sciencedirect-com-443.web.bisu.edu.cn/science/article/pii/S0301420718304938

27.

Töbelmann

Wendler

The impact of environmental innovation on carbon dioxide emissions. Journal Clean Prod [Internet]. 2020 [cited 2024 May 6];86:690–695. Available from: https://www-sciencedirect-com-443.web.bisu.edu.cn/science/article/pii/S0959652619336571

28.

Zhang

Luo

, et al. The effect of emission trading policy on carbon emission reduction: Evidence from an integrated study of pilot regions in China. Journal Clean Prod [Internet]. 2020 [cited 2024 May 6];53:321–330. Available from: https://www-sciencedirect-com-443.web.bisu.edu.cn/science/article/pii/S0959652620318904

29.

Wenger

Stauffacher

Dallo

. Public perception and acceptance of negative emission technologies–framing effects in Switzerland. Climatic Chang [Internet]. 2021 [cited 2025 Feb 17];167:932–940. Available from: https://link-springer-com-443.web.bisu.edu.cn/article/10.1007/s10584-021-03150-9

30.

Tangato

. The impact of clean technology adoption on carbon emissions: A global perspective. Clean Technol Environ Policy 2024; 78: 1043–1050.

31.

Linzenich

Engelmann

Arning

, et al. Harmful or beneficial to humans and the environment? An empirical study on the social acceptance and risk perception of CO2-based fuels. Front Environ Sci 2022; 10: 943–950.

32.

Khan

Hou

HP.

The impact of natural resources, energy consumption, and population growth on environmental quality: fresh evidence from the United States of America. Sci Total Environ [Internet]. 2021;754:142222. Available from: https://www-sciencedirect-com-443.web.bisu.edu.cn/science/article/abs/pii/S004896972035751X?via%3Dihub

33.

Liu

Khan

Haider

, et al. Financial development and environmental degradation: promoting low-carbon competitiveness in E7 Economies’ industries. Int J Environ Res Public Heal [Internet] 2022, [cited 2024 Nov 18]; 19: 16336. Available from: https://www.mdpi.com/1660-4601/19/23/16336/htm

34.

Moaniba

Does innovation respond to climate change? Empirical evidence from patents and greenhouse gas emissions. Technological Forecast Soc Chang [Internet]. 2017 [cited 2024 May 6];90:403–410. Available from: https://www-sciencedirect-com-443.web.bisu.edu.cn/science/article/pii/S0040162516302542

35.

Gotsch

Hipp

. Measurement of innovation activities in the knowledge-intensive services industry: a trademark approach. Serv Ind J 2012; 32: 2167–2184.

36.

Khan

Hamid

Sharif

, et al. Impact of adaptation technologies and green energy on environmental quality: Evidence from G8 economies. Gondwana Res [Internet]. 2025 [cited 2025 Feb 17];54:550–559. Available from: https://www-sciencedirect-com-443.web.bisu.edu.cn/science/article/pii/S1342937X24003502

37.

Liu

Ren

, et al. Do real output and renewable energy consumption affect CO2 emissions? Evidence for selected BRICS countries. 2020 [cited 2024 Apr 30]; Available from: https://www.mdpi.com/1996-1073/13/4/960

38.

Liu

Ren

Cheng

, et al. The role of globalization in CO2 emissions: a semi-parametric panel data analysis for G7. Science Total Environ [Internet]. 2020 [cited 2024 Apr 30];23:880–889. Available from: https://www-sciencedirect-com-443.web.bisu.edu.cn/science/article/pii/S0048969720308895

39.

Durani

Anwar

Sharif

, et al. Exploring the link between inflation, tourism, trade, foreign direct investment, and renewable energy consumption: evidence from novel wavelet quantile correlation approach. Clean Technol Environ Policy 2024; 16: 201–210.

40.

Ibrahiem

Hanafy

. Dynamic linkages amongst ecological footprints, fossil fuel energy consumption and globalization: an empirical analysis. Manag Environ Qual An Int J 2020; 31: 1549–1568.

41.

Saud

Chen

Haseeb

The role of financial development and globalization in the environment: accounting ecological footprint indicators for selected one-belt-one-road initiative countries. Journal Clean Prod [Internet]. 2020 [cited 2024 May 6];26:10065–10066. Available from: https://www-sciencedirect-com-443.web.bisu.edu.cn/science/article/pii/S0959652619343884

42.

Doytch

The impact of foreign direct investment on the ecological footprints of nations. Environmental Sustain Indic [Internet]. 2020 [cited 2024 May 6];74:20001–20009. Available from: https://www-sciencedirect-com-443.web.bisu.edu.cn/science/article/pii/S2665972720300696

43.

Georgatzi

Stamboulis

Vetsikas

Examining the determinants of CO2 emissions caused by the transport sector: empirical evidence from 12 European countries. Econ Anal Policy [Internet]. 2020 [cited 2024 Jun 7];65:11–20. Available from: https://www-sciencedirect-com-443.web.bisu.edu.cn/science/article/pii/S0313592619300232

44.

Sohag

Mariev

Davidson

. Revising environmental kuznets curve in Russian regions: role of environmental policy stringency. Environ Sci Pollut Res 2021; 28: 52873–52886.

45.

Wolde-Rufael

Mulat-Weldemeskel

. Do environmental taxes and environmental stringency policies reduce CO2 emissions? Evidence from 7 emerging economies. Environ Sci Pollut Res 2021; 28: 22392–22408.

46.

Bhowmik

Sharif

Anwar

, et al. Does environmental policy stringency alter the natural resources-emissions nexus? Evidence from G-7 countries. Geoscience Front [Internet]. 2024 [cited 2025 Feb 17];92:7211–7220. Available from: https://www-sciencedirect-com-443.web.bisu.edu.cn/science/article/pii/S1674987124000987

47.

Huang

, et al. Corporate social responsibility and bank credit loans: exploring the moderating effect of the institutional environment in China. Asia Pacific J Manag [Internet]. 2023 [cited 2025 Jun 3];40:707–742. Available from: https://link-springer-com-443.web.bisu.edu.cn/article/10.1007/s10490-021-09800-x

48.

Khaparde

Professor

The impact of corporate social responsibility on community well-being & social work intervention. Applied Ecology and Environmental Research [Internet] 2024 [cited 2025 Jun 3];2:351. Available from: www.jetnr.org

49.

Xuetong

Hussain

Rasool

, et al. Impact of corporate social responsibility on sustainable competitive advantages: the mediating role of corporate reputation. Environ Sci Pollut Res [Internet]. 2024 [cited 2025 Jun 3];31:46207–46220. Available from: https://link-springer-com-443.web.bisu.edu.cn/article/10.1007/s11356-023-28192-7

50.

Gazi

MAI

Hossain

Islam

, , et al. CSR And sustainable environmental performance: an exploration of mediating and moderating factors. Sustain [Internet]. 2024 [cited 2025 Jun 3];16:8499. Available from: https://www.mdpi.com/2071-1050/16/19/8499/htm

51.

Nogueira

Gomes

Lopes

JM.

Unveiling triple bottom line’s influence on business performance. Discov Sustain [Internet]. 2025 [cited 2025 Jun 3];6:1–18. Available from: https://link-springer-com-443.web.bisu.edu.cn/article/10.1007/s43621-025-00804-x

52.

Chilombe

Chiziwa

Chilombe

, et al. CSR and environmental responsibility eco-friendly practices. 2024 [cited 2025 Jun 3]; Available from: https://www.intechopen.com/chapters/1174286

53.

Vardhan

Alam

Raj

, et al. The impact of corporate social responsibility on corporate reputation. [cited 2025 Jun 3]; Available from: www.ijfmr.com

54.

Friede

Busch

Bassen

ESG And financial performance: aggregated evidence from more than 2000 empirical studies. Journal Sustain Financ Investment [Internet]. 2015 [cited 2025 Mar 5];5:210–233. Available from: https://www-tandfonline-com-s.web.bisu.edu.cn/doi/abs/10.1080/20430795.2015.1118917

55.

Nian

Said

FF.

The impact of ESG on firm risk and financial performance: a systematic literature review. J Scientometr Res [Internet]. 2024 [cited 2025 Jun 3]; 13:s144–s155. Available from: https://jscires.org/article/7878/

56.

Yang

Wang

Liu

ESG performance and corporate labor investment efficiency: Evidence from China. International Rev Econ Financ [Internet]. 2025 [cited 2025 Jun 3];56:5432–5440. Available from: https://www-sciencedirect-com-443.web.bisu.edu.cn/science/article/pii/S1059056025000723

57.

Grossman

Krueger

. Economic growth and the environment. Q J Econ 1995; 110: 353–377.

58.

WIPO. Patent analytics reports on environment and sustainability [Internet]. 2025 [cited 2025 Jun 3]. Available from: https://www.wipo.int/web/patent-analytics/environment-and-sustainability

59.

McBain

Lenzen

Albrecht

, et al. Building robust housing sector policy using the ecological footprint. Resour [Internet]. 2018 [cited 2025 Jun 3];7:24. Available from: https://www.mdpi.com/2079-9276/7/2/24/htm

60.

Režný

Bureš

Energy transition scenarios and their economic impacts in the extended neoclassical model of economic growth. Sustain [Internet]. 2019 [cited 2025 Jun 3];11:3644. Available from: https://www.mdpi.com/2071-1050/11/13/3644/htm

61.

Freeman

Strategic management: A stakeholder approach [Internet]. 2010 [cited 2025 Jun 3]. Available from: https://books.google.com/books?hl=en&lr=&id=NpmA_qEiOpkC&oi=fnd&pg=PR5&dq=Freeman,+R.E.+(1984)+Strategic+Management:+A+Stakeholder+Approach.+Pitm&ots=62foC5K8SO&sig=SGmdyg0KpvEk2o1XF2AO3-zFPNQ

62.

Kalra

. Corporate social responsibility and stakeholder theory: An integrated review. J Manag & Entrep [Internet]. 2024 [cited 2025 Jun 3];89:6209–6216. Available from: https://www.academia.edu/116768735/CORPORATE_SOCIAL_RESPONSIBILITY_AND_STAKEHOLDER_THEORY_AN_INTEGRATED_REVIEW

63.

Awa

Etim

Ogbonda

Stakeholders, stakeholder theory and corporate social responsibility (CSR). Int J Corp Soc Responsib [Internet]. 2024 [cited 2025 Jun 3];9:1–14. Available from: https://jcsr.springeropen.com/articles/10.1186/s40991-024-00094-y

64.

Batool

Mohsin

Alwadi

BM.

Green innovation and environmental performance: the moderating roles of governance and policy. World [Internet]. 2025 [cited 2025 Jun 3];6:29. Available from: https://www.mdpi.com/2673-4060/6/1/29/htm

65.

Broby

Smyth

On the use of principal components analysis in index construction. SSRN Electron J [Internet]. 2022 [cited 2025 Feb 17];90:810–816. Available from: https://papers.ssrn.com/abstract=4294718

66.

Swamy

Efficient inference in a random coefficient regression model. J Econom Soc [Internet]. 1970 [cited 2024 May 1];38:311–323. Available from: https://www-jstor-org-443.web.bisu.edu.cn/stable/1913012

67.

Hashem Pesaran

Yamagata

. Testing slope homogeneity in large panels. J Econom 2008; 142: 50–93.

68.

Pesaran

. A simple panel unit root test in the presence of cross-section dependence. Journal of Applied Econometrics 2007; 22: 265–312.

69.

Machado

JAF

Santos Silva

JMC.

Quantiles via moments. J Econom [Internet]. 2019;213:145–173. Available from:

70.

Lingyan

Zhao

Malik

, et al. Asymmetric impact of fiscal decentralization and environmental innovation on carbon emissions: evidence from highly decentralized countries. Energy Environ 2022; 33: 752–782.

71.

Guan

Hassan

Nassani

AA.

Investigating the role of environmental taxes, green finance, natural resources, human capital, and economic growth on environmental pollution using panel quantile regression [Internet]. 2025 [cited 2025 Feb 17]. Available from: https://www.mdpi.com/2071-1050/17/3/1094

72.

Dumitrescu

Hurlin

. Testing for granger non-causality in heterogeneous panels. Econ Model 2012; 29: 1450–1460.

73.

Westerlund

. Testing for error correction in panel data. Oxf Bull Econ Stat 2007; 69: 709–748.

74.

Herring

Sorrell

Energy efficiency and sustainable consumption. Springer, the rebound Eff Hampshire [Internet]. 2009 [cited 2025 Feb 17]; Available from: https://link-springer-com-443.web.bisu.edu.cn/content/pdf/10.1057/9780230583108.pdf