Economic development has led to significant improvements in human wellbeing, social welfare, and living standards across the globe. However, these gains often come at the expense of environmental sustainability, with scientists warning that continued ecological degradation may compromise the prospects of a sustainable life for future generations (Hunjra et al., 2023). Rapid industrialization, urbanization, and resource exploitation have intensified climate change, biodiversity loss, and ecosystem disruption. In response, policymakers and researchers have increasingly focused on the green economy as a comprehensive framework to reconcile economic growth with environmental sustainability.

Economic growth is widely recognized as a key driver of environmental degradation (Gorus and Aydin, 2019; Yi and Aziz, 2025), but its impact on ecological outcomes is not always linear. The Environmental Kuznets Curve (EKC) hypothesis posits a nonlinear relationship between income and environmental pressure, often taking an inverted U-shape: environmental degradation initially rises with economic growth, but after a certain income threshold, higher affluence allows for investment in cleaner technologies and more sustainable practices, reducing ecological impact (Al-Mulali et al., 2015; Dogan and Aslan, 2017; Jóźwik et al., 2025; Ulucak and Bilgili, 2018). Empirical studies provide mixed evidence, with some countries following the inverted U-shaped trajectory while others show U-shaped, N-shaped, or insignificant patterns depending on economic structure, policy frameworks, and technological advancement (Bekhet and Othman, 2018; Charfeddine, 2017; Mahmood et al., 2023; Sarkodie and Strezov, 2019; Destek and Sarkodie, 2019; Shahzad and Rahman, 2025). Considering EKC in the analysis highlights that economic development interacts with green factors, such as renewable energy, green finance, and capital formation, to shape the ecological footprint. Integrating these dimensions allows for a more nuanced understanding of how growth and sustainability objectives can coexist, particularly in developing countries.

A green economy is defined as a model that enhances human wellbeing and social equity while reducing ecological risks, promoting resource efficiency, and ensuring social inclusion (Voumik and Shah, 2014; Fankhauser et al., 2017). It is operationalized through strategies that improve environmental quality, facilitate sustainable production and consumption patterns, and promote good governance to maintain intergenerational welfare (Demiral and Demiral, 2021; Guarini et al., 2022; Mishra, 2017). The model emphasizes renewable energy adoption, sustainable infrastructure, and ecological conservation as core drivers of long-term economic resilience and environmental protection (Nguyen et al., 2024; Pratama et al., 2022).

Renewable energy represents a cornerstone of the green economy due to its capacity to reduce reliance on fossil fuels and mitigate greenhouse gas emissions. A substantial body of empirical evidence demonstrates that renewable energy consumption and production can reduce ecological footprints and improve environmental quality. For instance, Alola et al. (2019) find that renewable energy consumption significantly enhances environmental sustainability in 16 EU countries between 1997 and 2014, while Balsalobre-Lorente et al. (2018) corroborate similar effects in the EU-5 region during 1985–2016. Destek and Sinha (2020) highlight that renewable energy mitigates environmental degradation in 24 OECD countries, contrasting sharply with the negative environmental impacts of fossil energy. Evidence from BRICS countries (Ulucak and Khan, 2020) and Asian nations (Xu and Wu, 2023) further reinforces the environmental benefits of renewable energy. Renewable energy production is also critical, as it facilitates cleaner energy supply systems and technological innovation. Studies from Pakistan (Iqbal et al., 2022), Turkey (Bölük and Mert, 2015), and the EU (Alola et al., 2019) report that renewable energy production reduces CO2 emissions and ecological footprints in the long run, promoting cleaner industrial and transportation systems. These findings suggest that investments in renewable energy infrastructure and technology are essential for sustainable development.

Despite the generally positive consensus, some research highlights regional or contextual limitations. For instance, Sharif et al. (2020) find that renewable energy production in Turkey may inadvertently degrade environmental quality due to inefficient technologies, while studies by Çakmak and Acar (2022), Pata (2018), Sahbi and Shahbaz (2014), and Albayrak et al. (2022) report insignificant or even positive associations between renewable energy and CO2 emissions or ecological footprint in certain contexts. Sun (2023) underscores regional heterogeneity within Asia, showing that renewable energy reduces emissions in Eastern, Western, and Central Asia but has limited effects in Southeastern and Southern Asia. These discrepancies highlight that the environmental benefits of renewable energy are contingent on technological maturity, energy infrastructure, policy effectiveness, and regional socio-economic conditions. Moreover, renewable energy adoption is not only a technical challenge but also an institutional and financial one. The presence of incentives, subsidies, or supportive regulatory frameworks strongly shapes the scale and speed of renewable energy deployment, influencing its ultimate impact on ecological sustainability.

Green finance has emerged as a complementary mechanism to drive the transition toward a green economy by allocating financial resources toward environmentally sustainable projects. Green finance includes green bonds, green credits, environmental subsidies, and investment funds that prioritize renewable energy, green infrastructure, and sustainable production (Chen and Chen, 2021; Zhou and Li, 2022; Zakari, 2022). Its central role is to bridge the financing gap for green development, channelling resources from inefficient and polluting sectors to cleaner, low-carbon alternatives (UNEP, 2015). Empirical evidence indicates that green finance can significantly enhance environmental quality (Cao, 2023; Chen and Chen, 2021; Wan et al., 2022; Yi and Aziz, 2025). For example, in China, green finance policies have been shown to reduce carbon emissions, improve corporate environmental performance, and attract private capital toward renewable energy projects (Zhou et al., 2020; Zhang and Wang, 2021; Ren et al., 2020). Xu and Dong (2023) demonstrate that green investment promotes ecological integrity by financing clean energy and sustainable infrastructure. Hunjra et al. (2023) argue that green finance not only reallocates capital efficiently but also enhances the energy structure of the economy, thereby improving local quality of life.

Nevertheless, the efficacy of green finance is not uniform across contexts. Studies reveal that in developing or less developed countries, weak financial infrastructure and limited private sector engagement can constrain the effectiveness of green finance (Fu and Ng, 2021; Rasoulinezhad and Taghizadeh-Hesary, 2022). In the United States, Hammoudeh et al. (2020) find that the link between green bonds and environmental outcomes weakened after 2015, suggesting that initial gains may not always persist without supportive regulatory frameworks. Research from African nations (Zhang et al., 2023) and China (He et al., 2019) also shows that misaligned green financing can inadvertently slow renewable energy adoption or fail to reduce emissions. The literature indicates that green finance is most effective when complemented by sound governance, strong institutions, and public-private cooperation, highlighting the importance of integrating financial, technological, and policy dimensions in sustainable development strategies.

Capital formation represents another crucial determinant of ecological footprint. As a driver of economic production, capital formation influences resource consumption, energy use, and environmental pressures. Empirical studies consistently report a positive relationship between gross capital formation and CO2 emissions or ecological footprint. Gao et al. (2020) show that capital formation in China doubled CO2 emissions between 2007 and 2017. Similar findings are reported for OECD countries (Mujtaba et al., 2022), G20 countries (Li et al., 2023), Algeria (Chekouri et al., 2023), and the United States (Khan and Hou, 2021). These findings suggest that while capital accumulation is essential for economic growth, its environmental consequences are substantial unless investment is channelled toward sustainable sectors. Some studies, however, reveal nuanced effects. Rahman and Ahmad (2019) find that positive shocks in capital formation exacerbate CO2 emissions, whereas negative shocks may improve environmental quality. Bukhari et al. (2014) further confirm the long-term environmental impacts of capital accumulation in Pakistan, highlighting the need to align capital formation strategies with sustainability objectives. The literature suggests that green capital formation, i.e., investment directed toward renewable energy, energy efficiency, and pollution mitigation, can mitigate the adverse effects of traditional capital accumulation, providing an avenue for decoupling economic growth from environmental degradation.

Several recent studies have employed GMM-based approaches to examine ecological footprint dynamics, highlighting the importance of addressing endogeneity and persistence in environmental indicators (e.g., Al-Mulali et al., 2015; Çakmak and Acar, 2022). Our study builds on this literature by integrating green finance and renewable energy channels within a dynamic STIRPAT framework focused on developing economies. The STIRPAT (Stochastic Impacts by Regression on Population, Affluence, and Technology) framework extends the classical IPAT identity by allowing stochastic estimation of environmental impacts and flexible inclusion of additional socioeconomic and technological variables. In this framework, environmental pressure is modeled as a function of population dynamics (P), economic affluence (A), and technological and structural characteristics (T), while allowing for heterogeneous effects across countries and over time.

In empirical applications, STIRPAT is often extended to incorporate variables capturing energy systems, financial development, institutional quality, and technological change, particularly in cross-country analyses where environmental outcomes are shaped by multiple interacting channels. This flexibility makes the STIRPAT framework suitable for analyzing ecological footprint dynamics in developing economies, where demographic, economic, and technological transitions occur simultaneously.

Overall, the literature illustrates that ecological footprint is determined by the interplay of renewable energy adoption, green finance, and capital formation within the broader context of economic development. While renewable energy and green finance generally mitigate environmental degradation, their effectiveness is influenced by regional characteristics, institutional capacity, and the allocation of resources. Capital formation, if not guided toward green investments, may intensify environmental pressures, underscoring the need for integrated policy frameworks that simultaneously promote economic growth and ecological sustainability. Despite extensive research on individual factors, relatively few studies examine the combined effects of renewable energy production and consumption, green finance, and capital formation on ecological footprint, particularly in developing countries. Such integration is crucial to understand whether synergies exist among these factors and how they collectively influence environmental outcomes. Addressing this gap is essential for designing coherent strategies that leverage renewable energy, financial instruments, and capital allocation to achieve sustainable development while minimizing ecological impact. Future research should focus on multi-country, multi-period analyses to uncover dynamic interactions among these variables and guide evidence-based policy interventions.

Prior to estimation, the stationarity of the variables was examined using the Im–Pesaran–Shin (IPS) unit root test, which is suitable for unbalanced panel data and allows for missing observations. The results indicate that all variables are stationary either at level or at first difference, confirming that none are integrated of order two. This ensures that the regression estimates are consistent, and the analysis is free from spurious correlations.

The empirical investigation began with Ordinary Least Squares (OLS) estimation to provide a baseline understanding of the relationship between ecological footprint (EF) and the explanatory variables. The OLS model demonstrates good overall fit, with the F-statistic significant at nearly zero percent and an adjusted R ² of approximately 0.48, suggesting that almost half of the variation in EF is explained by the model. The t-statistics reveal that, except for government effectiveness (Gov), all explanatory variables exert statistically significant effects on EF. Specifically, green finance (GF), renewable energy consumption (REcons), remittances (Remit), and urbanization (Urban) significantly reduce EF at the 1% level, implying that these factors play a key role in promoting environmental sustainability. Conversely, renewable energy production (REpro), gross capital formation (GCF), trade openness (Trade), and financial development (FD) are positively associated with EF, indicating that capital accumulation and trade expansion may increase environmental pressure in the absence of strong environmental regulations or clean technologies. These findings are largely consistent with the pairwise correlation results, except for REpro, which displays a sign change in regression analysis.

To validate these initial estimates, several diagnostic tests were conducted. The Variance Inflation Factor (VIF) test confirms that multicollinearity is not a concern, and the Wooldridge test indicates no serial autocorrelation. However, the White test detects heteroskedasticity, suggesting that OLS estimates may not be fully efficient. Consequently, panel data techniques were employed to account for unobserved heterogeneity and improve estimation accuracy.

Fixed Effects (FEM) and Random Effects (REM) models were then estimated to account for country-specific and time-specific variations. Both models suggest the existence of unobserved heterogeneity, and the Hausman test strongly favors the Fixed Effects Model over the Random Effects Model, confirming that country-specific characteristics are correlated with the explanatory variables. However, the Modified Wald test indicates groupwise heteroskedasticity within the FEM, implying that cross-sectional heterogeneity persists. Although Generalized Least Squares (GLS) could correct for heteroskedasticity, it remains insufficient to address potential endogeneity issues inherent in dynamic panels.

To further test for endogeneity, the Instrumental Variable Two-Stage Least Squares (IV-2SLS) approach was applied. The results confirm that the first lag of EF is endogenous, indicating that current ecological outcomes depend on past environmental performance. Hence, the two-step System Generalized Method of Moments (System GMM) estimator was adopted to correct for endogeneity, autocorrelation, and heteroskedasticity simultaneously. This method is particularly suitable given that the dataset includes more cross-sectional units (countries) than time periods. Diagnostic statistics, including the Arellano–Bond AR (2) and Hansen tests, validate the absence of second-order serial correlation and confirm the validity of the chosen instruments. The number of instruments generated by the System GMM estimator is explicitly reported in all tables and remains lower than the number of countries, in line with Roodman (2009), ensuring that instrument proliferation is effectively controlled. Therefore, the System GMM estimates can be regarded as robust and reliable.

Regression results from all estimation methods (OLS, FEM, REM, GLS, IV-2SLS, and System GMM) are summarized in Table 4. The consistency in the direction of coefficients for the key variables (GF, REcons, REpro, and GCF) across estimation techniques reinforces the robustness of the results. Although the magnitude and statistical significance of coefficients vary slightly, the findings consistently show that green finance and renewable energy consumption significantly reduce EF, while renewable energy production and capital formation tend to increase it. These patterns underscore the complexity of the green transition process in developing countries, where financial, technological, and developmental dynamics interact in shaping ecological outcomes.

The empirical results align with much of the recent literature emphasizing the environmental benefits of renewable energy consumption. Studies such as Alola et al. (2019), Destek and Sinha (2020), and Balsalobre-Lorente et al. (2018) similarly find that renewable energy use reduces environmental degradation across EU and OECD countries. The negative and significant impact of green finance on EF is consistent with Zhou et al. (2020) and Xu and Dong (2023), who highlight that green financial flows encourage renewable energy investment and environmentally responsible projects. The positive association between capital formation and EF corroborates findings from Gao et al. (2020) and Mujtaba et al. (2022), indicating that conventional capital accumulation, when not directed toward green sectors, tends to intensify environmental pressure. Moreover, the positive effect of renewable energy production suggests that in developing countries, energy generation technologies may still rely on transitional or less efficient systems—consistent with Sharif et al. (2020) and Çakmak and Acar (2022), who note that the environmental benefits of renewable energy depend on technological maturity and institutional effectiveness.

Overall, these findings suggest that while financial and renewable energy mechanisms contribute to ecological sustainability, their effectiveness depends on complementary institutional frameworks and technological readiness.

Most independent variables exhibit consistent directions of impact on the ecological footprint (EF) across all estimation approaches, though a few variables show mixed behavior. In particular, renewable energy production (REpro) demonstrates some inconsistency: it displays a positive effect on EF in the OLS model but a negative (though insignificant) effect under the Fixed Effects Model (FEM). However, in the GLS and System GMM estimations, REpro exerts a positive and statistically significant impact at the 5% level. This suggests that, while renewable electricity generation generally contributes to environmental improvement in theory, in practice, production inefficiencies and transitional energy technologies in developing countries may offset its expected environmental benefits.

Conversely, industrialization (Indus) shifts from a positive correlation with EF in the pairwise matrix to a negative and statistically significant coefficient in the OLS and GLS regressions. In the System GMM estimation, Indus remains positively but statistically insignificant, indicating that its independent contribution to EF is relatively weak once other economic and green factors are accounted for. These results imply that the effects of renewable energy production and industrialization on environmental outcomes are context-dependent and may vary across countries and development stages.

For the remaining explanatory variables, the direction and significance of coefficients are remarkably stable across all models (OLS, FEM, REM, GLS, and GMM) confirming the robustness of their relationships with EF. In particular, green finance (GF), renewable energy consumption (REcons), remittances (Remit), and urbanization (Urban) consistently exert negative and significant impacts on EF, suggesting that financial support for renewable projects, cleaner energy use, and remittance-driven income flows collectively contribute to reducing ecological pressure. In contrast, gross capital formation (GCF), renewable energy production (REpro), and trade openness (Trade) show positive and statistically significant effects on EF across multiple estimators, indicating that economic expansion and trade activities still rely heavily on energy-intensive and resource-demanding structures. Nevertheless, given the dynamic panel structure, diagnostic validity rather than coefficient stability remains central, which is why formal GMM diagnostics and instrument controls are emphasized.

These findings reinforce the reliability of the System GMM results as the preferred and most efficient estimation for this panel dataset. The consistency of core coefficients across different estimation techniques strengthens the empirical evidence for policy implications: promoting green finance and renewable energy consumption can significantly mitigate ecological footprint, while unchecked capital accumulation and trade expansion may intensify environmental pressures unless guided by green and efficient technologies.

To enhance the robustness of the results, the GMM approach was applied with the inclusion of additional macroeconomic variables one at a time (Table 5). All diagnostic tests, including the number of instruments, group balance, over-identification restrictions, and autocorrelation (assessed via the Hansen test and Arellano–Bond test), satisfied the standard requirements, ensuring the model’s reliability and validity. The results remain consistent, confirming the directional impact of green factors and other variables, with only minor changes in coefficient magnitudes. Notably, the first lag of the dependent variable exhibits the largest coefficient, significant at the 1% level. This indicates a strong persistence of the ecological footprint (EF), implying that environmental degradation in the previous period significantly influences the current environmental situation.

Green finance, represented by public investment in renewable energy, and green economic activities, proxied by renewable energy consumption, exert a negative effect on EF, significant at the 10% level. This suggests that investments in renewable energy and policies promoting its consumption are effective measures to mitigate EF, which is critically important for environmental quality and the sustainability of natural resources. These findings align with the conclusions of Hunjra et al. (2023), Sun (2023), Mehboob et al. (2024), and Sun et al. (2023), confirming the role of green finance and renewable energy consumption in environmental protection. Consequently, increasing green investments and replacing fossil fuels with renewable energy sources represent viable strategies to improve environmental quality while supporting sustainable economic development.

The negative coefficient of remittances (REMIT) indicates that remittances contribute positively to environmental improvement. This finding is consistent with Wang et al. (2021), who analyzed five remittance-receiving countries from 1980-2016, but contrasts with Yang et al. (2020) and Jamil et al. (2022), who studied 97 countries (1990–2016) and the G20 period (1990–2019), respectively. This suggests that the role of remittances in environmental protection may be context-specific, particularly for developing countries, which predominantly receive remittances. Remittances can enhance household wellbeing and purchasing power, and if allocated toward eco-friendly investments and consumption, they may serve as a valuable capital source for environmental protection.

Similarly, urbanization exerts a negative effect on EF in the current analysis. This aligns with findings by Ulucak and Khan (2020) for BRICS economies (1992–2016) and Sharma (2011) for a sample of 69 countries. Urbanization initially increases energy demand and reliance on cheap fossil fuels, which can elevate greenhouse gas emissions and impact public health. However, higher income and purchasing power can raise environmental awareness, prompting demand for cleaner consumption and production patterns, supported by regulations integrating environmental protection into economic activities. Empirical evidence also highlights the positive role of green cities in environmental quality, with high-income countries achieving notable environmental outcomes through sustainable urbanization (Charfeddine, 2017; Liu and Bae, 2018). Nonetheless, some studies report ambiguous effects of urbanization on EF (Sadorsky, 2014; Chikaraishi et al., 2015), cautioning policymakers regarding its complex interactions. Bargaoui et al. (2014) suggest that urbanization may harm environmental quality in middle-income countries while reducing CO₂ emissions in high-income countries. The negative impact of urbanization on EF observed here indicates that developing countries may pursue urban development while maintaining environmental sustainability, particularly through green investment and clean energy initiatives.

For variables exerting a positive impact, gross capital formation (GCF), renewable energy production (REpro), and trade openness significantly increase EF at the 5% level. Capital formation stimulates energy and resource demand, contributing to environmental degradation through the scale effect (Jóźwik et al., 2025). While investment is essential for economic growth and production capacity, it often prioritizes low-cost resources, which can elevate greenhouse gas emissions. The positive relationship between GCF and EF indicates that investments in environmentally harmful sectors remain prevalent, highlighting the need for policies guiding capital toward eco-friendly activities.

Regarding renewable energy production, the positive coefficient of REpro suggests that, in the short term, increased production of renewable energy coincides with higher EF. This observation is consistent with Albayrak et al. (2022), Chien et al. (2023) and Sharif et al. (2020) and can be attributed to the industrialization process in developing countries, which heavily relies on cheap resource extraction, fossil fuel consumption, and energy-intensive industries. However, we note here that the positive association between renewable energy production and ecological footprint should not be interpreted as evidence that renewable energy is inherently environmentally harmful. Instead, this result reflects transitional dynamics in developing economies, where renewable energy expansion often occurs alongside fossil-based systems, energy-intensive infrastructure development, and industrial scale effects. During early stages of technological transition, increases in renewable energy production may coincide with higher material use, land conversion, and energy demand, temporarily increasing ecological pressure. Moreover, renewable energy production and renewable energy consumption are closely linked within national energy systems, yet they capture different dimensions of the energy transition. Production reflects supply-side capacity expansion and infrastructure investment, while consumption reflects demand-side substitution away from fossil fuels. In developing economies, these processes may evolve at different speeds, leading to distinct environmental effects in the short and medium term.

Although renewable energy constitutes a cleaner source, its share in total energy consumption remains relatively low (approximately 30%). Consequently, fossil fuel consumption continues to rise to meet industrial demand, resulting in higher EF. If renewable energy production can grow faster than total energy demand, fossil fuel consumption and greenhouse gas emissions may decline. Moreover, the larger negative coefficient of REcons relative to REpro confirms the net positive effect of renewable energy on environmental quality, suggesting that policies encouraging renewable energy consumption are crucial.

Trade openness (Trade) also exerts a positive effect on EF. Wu (2022) demonstrates that trade liberalization can exacerbate natural resource depletion in developing countries. This outcome can be explained through the pollution haven and composition effects (Abid, 2016). Trade encourages countries to specialize according to comparative advantage, with high-income countries leveraging advanced technology and stricter environmental regulations to engage in cleaner production, while importing resource-intensive goods from developing countries. Conversely, developing countries often prioritize economic growth over environmental protection, focusing on energy-intensive, low-tech industries for export. Consequently, increased trade can lead to environmental degradation in developing countries due to the composition effect. Limited access to finance, patents, and green technologies further constrains these countries’ ability to adopt environmentally friendly practices. To mitigate the negative environmental impacts of trade, developing countries require support in accessing advanced technologies, promoting green investments, improving resource efficiency, and transitioning toward sustainable, eco-friendly economies.

To examine differences across income levels, four dummy variables were created based on the World Bank classification: (1) high income (HighInc), (2) upper-middle income (UppMidInc), (3) lower-middle income (LowMidInc), and (4) low income (LowInc). The regression results for income groups indicate that all dummy variables are significant at the 5% level, with coefficients increasing from lower to higher income groups. Other explanatory variables retain the same positive or negative effects on EF as in the original regression model (Table 6).

When regressions are conducted separately for each income group, the income dummy variable becomes insignificant; however, all explanatory variables remain statistically significant and maintain their effect direction on EF. This confirms the robustness of the results regarding the roles of explanatory variables.

Ecological issues are universal, affecting all countries regardless of income level. While the impact of green factors may differ across income groups, coordinated international policy efforts are essential to address environmental challenges. The larger positive coefficients for high-income groups suggest that these countries need to make more substantial efforts to reduce EF.

Six regional dummy variables were created to analyze regional differences: East Asia and Pacific (region_dummy1), Europe and Central Asia (region_dummy2), Latin America and Caribbean (region_dummy3), Middle East and North Africa (region_dummy4), South Asia (region_dummy5), and Sub-Saharan Africa (region_dummy6). The regression results are presented in Table 7.

Similar to the income group analysis, in the aggregate model, all region dummy variables are significant at 5%. When regressions are run individually for each region, the region dummy variable is mostly insignificant, except for Latin America and the Caribbean (region_dummy3). All explanatory variables remain significant, confirming the robustness of the results. GF and REcons are particularly important for reducing EF.

These results emphasize the importance of regional and international cooperation in implementing environmental policies. Developing countries in Latin America and the Caribbean (region 3) and Europe and Central Asia (region 2) may need to make additional efforts to address EF compared to other regions.