Unveiling the dynamics of renewable and non-renewable energy consumption in driving environmental sustainability in sub-Saharan Africa

1 Introduction

Environmental sustainability has become one of the defining challenges of the twenty-first century, driven by the urgent need to balance economic development with the preservation of natural ecosystems. Rapid industrialization, population growth, and the intensive use of fossil fuels have contributed to widespread environmental degradation, with carbon dioxide (CO₂) emissions emerging as a leading driver of global warming (Nathaniel et al., 2020). Sustainability, however, goes beyond the avoidance of harm; it involves the careful management of natural resources to ensure their availability for future generations (Barbier, 1987). At the center of this discourse lies energy. While energy is indispensable for economic progress, the way it is produced and consumed strongly shapes environmental outcomes (Stern, 2004).

Globally, the energy sector remains a primary contributor to greenhouse gas (GHG) emissions. Fossil fuel consumption, particularly in industrializing economies, continues to accelerate climate change and ecological degradation (Nathaniel et al., 2020). Recognizing these challenges, international frameworks such as the Paris Agreement have sought to reduce emissions while promoting sustainable development pathways (International Energy Agency, 2022). Likewise, the evolution from the Millennium Development Goals to the Sustainable Development Goals (SDGs) placed greater emphasis on the interdependence between energy use, environmental protection, and human wellbeing. In particular, SDG 7 calls for universal access to affordable, reliable, and sustainable energy, while SDGs 12, 13, 14, and 15 reinforce the need for sustainable production and consumption, climate action, and the conservation of ecosystems (Morton et al., 2017).

Despite technological progress and renewable energy adoption in high-income countries, the transition to clean energy has been much slower in developing regions. Sub-Saharan Africa (SSA) illustrates this paradox. The region contributes less than 4% to global CO₂ emissions yet faces rising energy demands due to industrialization, rapid urbanization, and population growth (Adebayo et al., 2022; International Energy Agency, 2022). More than half of the region's population still lacks access to electricity, and millions continue to rely on traditional biomass such as firewood and charcoal for cooking and heating (Qureshi et al., 2019). This dependence on inefficient energy sources contributes not only to environmental degradation through deforestation, soil erosion, and rising emissions, but also to severe health problems linked to indoor air pollution (Ite and Ite, 2024). Africa's abundant natural resources further complicate this dynamic. The continent holds over 30% of the world's mineral reserves, 8% of global oil reserves, and 7% of natural gas reserves. Mineral trade alone accounts for about 28% of Africa's GDP and 70% of its total exports, serving as the largest source of public revenue in many countries (African Development Bank, 2023). The extraction of natural resources could generate more than USD 30 billion in annual government revenue over the next two decades, while new oil, gas, and mineral discoveries may increase government revenue by 9% to 31% within a decade of exploitation in countries such as Ghana, Mozambique, and Tanzania (African Development Bank, 2023). Beyond extractives, Africa's vast forests, water basins, and national parks provide enormous potential for ecotourism and renewable energy development. Yet, large-scale deforestation, mining, and oil extraction continue to accelerate environmental degradation (Schuhen, 2021).

In SSA, the intersection of industrialization and energy consumption is particularly significant. Industrialization is central to economic transformation, but it remains energy-intensive and heavily reliant on fossil fuels (Jermsittiparsert, 2021). As energy demand rises, the shift from biomass toward modern energy sources, including electricity and natural gas, is gaining momentum. However, this transition remains dominated by non-renewable energy, which undermines sustainability (Lv, 2023). Countries such as Kenya, South Africa, and Morocco have demonstrated the potential of solar, hydro, and wind power to diversify the energy mix (non-renewable and renewable energy), yet renewable energy adoption across SSA remains far below its potential (Adebayo et al., 2022). This creates a paradox: while energy consumption is vital for growth and poverty alleviation, it is also a key driver of CO₂ emissions, biodiversity loss, and ecosystem degradation (Kahia and Charfeddine, 2019). These realities highlight a critical research gap. Although the relationship between energy, industrialization, and the environment has been widely studied at the global and regional levels, the specific dynamics of SSA remain underexplored. The region faces unique challenges in energy access, infrastructure development, and reliance on traditional fuels (Nwala et al., 2023). A closer examination of how industrialization and the energy mix interact to shape environmental outcomes in SSA is therefore essential for evidence-based policymaking.

In this context, the unique added value of this study lies in its integrated and SSA-specific perspective on environmental sustainability. Contrary to most studies in the extant literature that explore industrialization, energy consumption, or environmental outcomes in isolation, this paper explicitly analyzes the roles of industrialization, renewable, and non-renewable energy in shaping environmental sustainability in SSA. By jointly accounting for fossil fuel dependence and renewable energy penetration in the industrialization process, the study captures the region's structural energy constraints and sustainable development realities. This approach provides policy-relevant evidence on how alternative energy pathways can moderate the environmental costs of industrial growth, thereby offering clearer guidance for SSA countries seeking to reconcile economic transformation with long-term environmental sustainability. The rest of this paper is structured as follows: Section 2 presents a review of the related literature. Section 3 discusses the methodological framework, while Section 4 analyzes the empirical findings and situates them within existing debates. Finally, Section 5 concludes the paper and offers policy recommendations for promoting environmental sustainability in SSA.

2 Review of literature

This section presents a comprehensive review of relevant literature to provide a solid foundation for the study. It examines the theoretical underpinnings and previous empirical findings related to the research topic.

2.1 Theoretical literature

The environmental Kuznets curve (EKC) hypothesis, introduced by Kuznets (1955) and later adapted by Grossman and Krueger (1991, 1995), provides one of the earliest frameworks for understanding the growth–environment nexus. It suggests an inverted-U relationship in which environmental degradation rises during the early stages of economic growth, driven by industrialization, energy consumption, and urbanization, but eventually declines as societies achieve higher income levels and begin investing in cleaner technologies, stricter regulations, and sustainable practices. Central to the EKC is the concept of a “turning point,” the stage at which economic progress begins to improve rather than harm the environment. Although the theory presents a hopeful narrative that growth can ultimately support sustainability, critics argue that it does not universally apply across countries or pollutants and that it overlooks complex issues such as biodiversity loss and climate change.

Building on these debates, ecological modernization theory (EMT) emerged in the 1980s and was formalized by Mol and Spaargaren (2000) as a more optimistic framework. EMT posits that economic growth and environmental protection can be compatible when societies embrace technological innovation, market mechanisms, and policy reforms. Unlike the EKC, which describes a long-term income-driven dynamic, EMT emphasizes deliberate actions such as the adoption of clean technologies, the integration of renewable energy, and the introduction of regulatory reforms that can immediately steer economies toward sustainability. EMT has shaped environmental policies worldwide, yet it is sometimes criticized for its heavy reliance on technological and market-based solutions, which may neglect the social and political dimensions of sustainability transitions.

In contrast to the optimism of EMT, the pollution haven hypothesis offers a more critical view of globalization's environmental consequences. Emerging in the 1980s and popularized by Copeland and Taylor (1994), it argues that firms in pollution-intensive industries often relocate to countries with weaker environmental regulations in order to reduce compliance costs. This process creates “havens” for dirty production. While empirical evidence is mixed, since other factors such as labor costs, infrastructure, and market access often weigh more heavily in firm location decisions, the hypothesis highlights the uneven environmental burden borne by developing countries. This dynamic raise important concerns about global equity and the sustainability of trade liberalization in the absence of harmonized environmental standards.

Recognizing the limitations of these perspectives, sustainable transition theory (Geels et al., 2004) shifts attention toward the structural transformations necessary for achieving sustainability. It emphasizes multi-level dynamics of change, ranging from innovations at the niche level such as solar, wind, and hydropower technologies to their interaction with established socio-technical regimes and the influence of broader landscape pressures such as globalization, climate change, and demographic shifts. Unlike the EKC's reliance on income growth or EMT's focus on technological and market reforms, transition theory underscores the need for systemic change involving governance, social practices, and collaboration among multiple stakeholders. This framework is particularly relevant for regions such as SSA, where sustainability challenges require not only cleaner technologies but also fundamental shifts in energy systems, institutions, and policy frameworks.

2.2 Empirical literature

The empirical review surveys previous studies that have examined similar research problems or related variables, highlighting the methodologies employed, key findings, and the contexts in which the studies were conducted. For instance, Diallo and Ouoba (2024) assessed the influence of renewable energy consumption on environmental quality in SSA using the refactored instrumental variables method. Their study focused on variables such as renewable energy consumption and CO₂ emissions, analyzing data from 2000 to 2022. The findings revealed that renewable energy consumption significantly enhances environmental quality by reducing CO₂ emissions. The authors further emphasized that countries with higher shares of renewable energy in their energy mix, such as Kenya and Ethiopia, have experienced notable improvements in environmental sustainability. However, the study also highlighted persistent challenges, including inadequate infrastructure, high initial investment costs, and limited access to financing for renewable energy projects. Consequently, they concluded that scaling up renewable energy adoption in SSA requires targeted policies, such as subsidies for green energy technologies and international partnerships to support financing and capacity building, thereby underscoring the critical role of renewable energy in achieving the region's sustainability goals. In a different context, Zuhal and Göcen (2024) investigated the link between renewable energy use, CO₂ emissions, and economic growth in the United States from 1973:M01 to 2022:M06. Using both symmetric and asymmetric spectral Granger causality analysis, they found evidence of bidirectional causality among CO₂ emissions, renewable energy consumption, and economic growth. The asymmetric results further revealed two-way causal relationships between positive and negative shocks in these variables, indicating that renewable energy consumption plays a pivotal role in promoting sustainable economic growth while improving environmental quality in the United States. Extending the evidence to Saudi Arabia, Ben-Ahmed et al. (2025) examined the nexus between CO₂ emissions, renewable energy consumption (REC), and non-renewable energy consumption (NREC) over the period 1990–2019. The study employed the augmented Dickey–Fuller (ADF) and Phillips–Perron (PP) tests to assess stationarity, which revealed a mixed order of integration and justified the use of the autoregressive distributed lag (ARDL) model with CO₂ emissions as the dependent variable. Lag selection criteria were applied to establish cointegration, and causality analysis was conducted within the ARDL framework, supported by robustness checks. Their results confirmed a significant long-run relationship between CO₂ emissions, REC, and NREC. Specifically, a 1% increase in REC leads to a 0.21% reduction in CO₂ emissions, whereas a 1% rise in NREC corresponds to a 53.4% decrease in emissions. Based on these findings, the authors recommended policy measures aligned with Saudi Arabia's Vision 2030 to support a sustainable energy transition.

Several empirical studies have examined the nexus between energy consumption, economic growth, and environmental sustainability across different countries and contexts. For example, Akadiri et al. (2019) investigated the relationship among carbon emissions, electricity consumption, economic growth, and globalization in Turkey from 1970 to 2014. Using a multivariate framework and several econometric techniques, including the (Bayer and Hanck, 2013) cointegration test, the ARDL bounds testing approach, short- and long-run ARDL estimations, and the Toda–Yamamoto Granger causality test, the study found that electricity consumption and economic growth are the primary drivers of pollution, mainly due to fossil fuel dependence, whereas globalization had no significant impact. Based on these results, the authors recommended restructuring fossil fuel-based energy systems, prioritizing renewable energy, promoting electric vehicles as substitutes for conventional ones, and strengthening environmentally sensitive taxes. In the same research context, Apinran et al. (2022) examined the effect of electricity consumption on economic growth in Nigeria, controlling for labor, capital, and carbon emissions from 1981 to 2019. Using ARDL and dynamic ARDL (DYNARDL) simulations, they found that electricity consumption, labor, and capital positively but inelastically drive growth, while carbon emissions exert a significant negative effect. The study highlights the importance of a reliable electricity supply in enhancing economic growth and social stability, emphasizing the need for efficient labor, adequate financing, and regular replacement of outdated equipment. Additionally, the authors recommend policies that promote a green economy to support sustainable growth and environmental protection in Nigeria.

In India, Ahmad et al. (2023) analyzed the relationship between environmental degradation, technological innovation, and electricity consumption from 1981 to 2018 using time series data. Employing ARDL, fully modified OLS (FMOLS), dynamic OLS (DOLS), and Granger causality through a vector error correction model (VECM), the study found that urbanization, financial development, and technological innovation reduce carbon emissions and improve environmental quality in the long term, whereas economic growth and electricity consumption contribute to environmental degradation. The authors emphasize that promoting renewable energy is essential for reducing environmental harm without compromising economic growth. In China, Abbasi et al. (2022) explored the impact of fossil fuel energy, renewable energy, and GDP on carbon dioxide emissions from 1980 to 2018 using DYNARDL simulations and frequency domain causality (FDC) models. They found that fossil fuel consumption significantly increases CO₂ emissions in both the short and long term, while GDP raises emissions in the long term but mitigates them in the short term. Renewable energy contributes to reducing emissions in the short term, and the FDC analysis confirmed causality across all time horizons. The study underscores the importance of renewable energy for environmental sustainability and recommends long-term strategies to curb carbon emissions. Hou et al. (2023) analyzed environmental factors affecting carbon emissions in OECD countries from 1990 to 2019, focusing on fossil fuel energy, renewable energy, and GDP. By estimating separate models for fossil fuel and renewable energy and applying cointegration tests along with the CS-ARDL estimator, the study revealed that fossil fuel consumption and GDP increase carbon emissions, whereas renewable energy reduces them and promotes sustainable development. The findings suggest that economic growth can help lower emissions when combined with cleaner energy adoption, highlighting the pivotal role of renewable energy in achieving environmental sustainability. Additionally, Achuo et al. (2022) examined the effects of renewable and non-renewable energy consumption on environmental sustainability using a global panel of 173 countries from 1996 to 2020. Utilizing a fixed-effects regression model with Driscoll–Kraay robust standard errors, the study found that renewable energy reduces GHG emissions and enhances environmental quality, whereas non-renewable energy increases emissions. According to these researchers, the positive effect of renewable energy is more pronounced in developed countries, while the negative impact of non-renewable energy is stronger in developing nations. Their finding also highlights the critical role of natural resource endowments and institutional quality in promoting environmental sustainability. It equally emphasized the need for strong institutions in order to ensure efficient and sustainable energy utilization, particularly of renewable sources.

Kadria et al. (2022) show that in MENA net oil importing countries, strong environmental policies and effective governance encourage renewable energy adoption, while lobbying pressures hinder it, and higher renewable energy use improves public health. Güney (2019) finds that renewable energy significantly enhances sustainable development across 40 developed and 73 developing countries, measured by adjusted net savings, whereas non-renewable energy has a weaker effect. Alola et al. (2023) show that, in India, both renewable energy use and non-renewable energy efficiency improve environmental sustainability, though financial development and trade may reduce it. Compared to these studies, the present paper focuses on SSA, using a dynamic panel approach that captures persistence and long-term adjustments in environmental outcomes and relies on direct ecological indicators such as ecological footprint and GHG emissions. While previous work emphasizes governance, policy, or income effects, this study examines systemic energy transitions and their direct impact on environmental sustainability in a region with unique structural and institutional challenges. Despite the growing recognition that renewable energy can reduce resource depletion and environmental pressure, empirical evidence from SSA remains scarce. This gap is particularly relevant given SDG 7, which calls for a substantial increase in renewable energy's share of the global energy mix by 2030 and stronger international cooperation and investment in clean energy. By using comprehensive sustainability indicators, this study provides region-specific evidence on the role of renewable energy in advancing sustainable development in SSA. A summary of the empirical literature is provided in Table 1.

Table 1. Summary of empirical literature.

The symbols represent increasing and decreasing relationship between the dependent and independent variables.

3 Data source and methodology

3.1 Data source

This study employs a robust panel data methodology to investigate the impact of electricity consumption, fossil fuel consumption, and REC on environmental sustainability in SSA. All variables are obtained from reputable databases that provide high-quality, internationally comparable data. The dependent variables, ecological footprint (EFP) and GHGs, serve as proxies for environmental sustainability. EFP is measured in global hectares per capita, sourced from the Global Footprint Network (GFN), while GHGs are measured in metric tons of CO₂ equivalent per capita, obtained from the World Development Indicators of the World Bank. The EFP reflects the biological capacity of ecosystems relative to human demand and is widely recognized in sustainability research, whereas GHG captures the volume of emissions contributing to climate change and environmental degradation (Magazzino et al., 2025). In this study, we use both ecological footprint (EFP) and GHG emissions to offer a more complete picture of environmental sustainability. While GHG emissions focus specifically on the amount of carbon dioxide released into the atmosphere and its contribution to climate change, EFP takes a broader view, accounting for the total environmental pressure from human consumption, including land, water, food, and other resources, alongside carbon dioxide. Using the two proxies also helps as a robustness check and in the interpretation of the results, as actions that reduce GHG emissions do not always lead to a lower ecological footprint, and vice versa. Previous research supports this approach, highlighting that EFP provides a more integrative measure of human environmental impact, complementing the more targeted GHG metric (Çelekli and Zariç, 2023; Faizi et al., 2025).

The key independent variables include electricity consumption (kilowatt-hours per capita), fossil fuel consumption (percentage of total energy use), and REC (percentage of total final energy consumption), all sourced from the World Bank's World Development Indicators (WDI). The control variables include GDP per capita (constant 2015 US dollars), population growth (population growth rate), and industrialization (industry value added as a percentage of GDP), all obtained from the WDI to account for macroeconomic and structural factors that may influence the energy-environment nexus. The combination of the independent and control variables, as presented in Table 2, provides a robust empirical foundation for the analysis, while the descriptive statistics are reported in Appendix A1.

Table 2. Description of variables, measures, and expected signs.

Figures 1, 2 show the scatter plots for the dependent variables, ecological footprint (EFP), and GHGs in relation to the independent variables (REC, GDP, POP, IND, FFC, and EC). A similar pattern emerges in both figures: REC, population growth (POP), and industrialization (IND) display negative associations with the two measures of environmental sustainability, while GDP per capita, fossil fuel consumption (FFC), and electricity consumption (EC) exhibit positive associations. It is important to note that the observed negative relationship between the independent variables and the environmental sustainability indicators does not imply causation. Scatter plots only provide a preliminary visual indication of potential relationships. Determining whether these associations are statistically meaningful and causal requires formal empirical testing using advanced econometric methods, such as the system generalized method of moments (GMM) estimators applied in this study.

As illustrated in Figures A1, A2 in Appendix, South Africa, Mauritius, Botswana, and Gabon are the leading emitters of environmental degradation in Africa, whereas Rwanda, the Democratic Republic of Congo, Mozambique, and Madagascar exhibit relatively low levels of both ecological footprint and GHG emissions. Regarding renewable energy infrastructure, the Democratic Republic of Congo, Uganda, and Rwanda are the top performers, while South Africa, Mauritius, Chad, and Botswana rank lower in terms of REC. Similarly, South Africa leads the continent in fossil fuel and EC, highlighting its high reliance on non-renewable and polluting energy sources, as depicted in Figures A3, A4 in Appendix. In terms of GHG emissions, Africa as a whole register the lowest levels globally, although it exhibits a steady upward trend compared to other regions. The global pattern indicates that North America and Europe contribute the most to environmental degradation, but both regions show a declining trend over time, as shown in Figure A5 in Appendix. This underscores the critical importance of encouraging the construction and expansion of renewable energy infrastructures in SSA, enabling the region to transition toward cleaner energy sources and achieve long-term environmental sustainability.

3.2 Model and estimation techniques

Given the dynamic and complex relationship between energy consumption and environmental sustainability, this study adopts the GMM for estimation. GMM is particularly appropriate for addressing endogeneity and unobserved heterogeneity that are common in macro-panel data. This is critical because energy consumption, economic growth, and environmental degradation are likely to influence one another simultaneously. The dynamic panel GMM technique, originally developed by Arellano and Bond (1991) and later refined by Arellano and Bover (1995) and Blundell and Bond (1998), offers a sophisticated solution to the limitations of traditional panel estimators such as fixed effects and random effects.

System GMM is preferred over difference GMM due to its superior performance when variables display persistence over time. Unlike difference GMM, which relies solely on lagged levels as instruments, system GMM combines equations in both differences and levels, using lagged differences as instruments for the level equation and vice versa. This approach enhances the efficiency and reliability of estimates, particularly in panels with a small-time dimension (T) and a large cross-sectional dimension (N), as is the case for SSA countries observed over two decades.

The estimator is also useful for several reasons. First, it controls for unobserved country-specific effects through the first-difference transformation. Second, it addresses potential endogeneity of the explanatory variables by employing higher-order lags of regressors as valid instruments. Third, it mitigates issues of simultaneity and reverse causality, which are highly relevant in the nexus between energy, economic activity, and environmental outcomes. Additionally, GMM allows for the inclusion of lagged dependent variables as regressors, which is crucial for capturing the dynamic behavior of environmental sustainability indicators such as the ecological footprint.

The reliability of the estimator depends on key diagnostic checks. The Arellano–Bond test for autocorrelation ensures that the error term does not exhibit first-order serial correlation, while the absence of second-order autocorrelation AR(2) validates the specification. Instrument validity is assessed using the Hansen J-test, and to avoid instrument proliferation, the number of lags is restricted and instruments are collapsed in line with best practice (Roodman, 2009). Importantly, the total number of instruments is kept lower than the number of countries to maintain robustness.

Finally, the analytical model is anchored in the EKC hypothesis (Grossman and Krueger, 1995) and the STIRPAT framework (Wei, 2011). These theoretical frameworks justify the inclusion of economic growth, industrialization (technology), and energy consumption as key determinants of environmental sustainability. Meanwhile, the analysis begins with an exploration of the scatter plots, followed by the estimation of the GMM models. Two core models are estimated using the GMM, as outlined earlier. The first Equation 1 specifies the model with the ecological footprint (EFP) as the dependent variable, while the second Equation 2 considers GHG as the dependent variable. Both EFP and GHG serve as proxies for environmental sustainability in this study. Due to data constraints across SSA countries, the sample is limited to 27 countries.

It is assumed that renewable energy is an endogenous variable, and the estimation of the model without accounting for the possible endogeneity may result to bias estimates. Therefore, the study adopts a system GMM estimation technique for panel regression analysis with smaller time (T) to capture the effect of renewable energy and non-renewable energy on environmental sustainability among SSA member countries. The GMM estimation technique was developed by Blundell and Bond (1998). It is particularly useful in addressing the issue of endogeneity that may arise when estimating a dynamic panel in the presence of unobserved heterogeneity or correlated individual effect. It addresses the issue of endogeneity by instrumenting for the potential endogenous variable. Estimation of the effect of renewable energy on environmental sustainability without accounting for endogeneity will result in underestimation of the model if the estimate of the lagged environmental sustainability variable lies above the estimate of the polled ordinary least squared. By instrumenting the endogenous independent variables, we are assuming the error term is not correlated with the endogenous variables. The system GMM equations are therefore presented as follows:

EFPit=α1EFPit-1+ α2RECit+α3FFCit+α4ECit+α5Xit

+μi+δi+εit,(1)

GHGit=α1GHGit-1+ α2RECit+α3FFCit+α4ECit+α5Xit

+μi+δi+εit.(2)

Here, EFP and GHG represent the ecological footprint and GHG emissions, respectively, while; EFP_*it*−1and GHG_*it*−1 denote the lagged values of the dependent variables. REC measures renewable energy consumption, FFC refers to fossil fuel production, and EC captures electricity consumption, while X_*it* is a vector of control variables that includes economic growth, industrialization, and population growth. The term μ_*i* accounts for unobserved country-specific effects, while δ_*i*captures the time trend. The coefficients α₁, α₂, α₃, α₄, and α₅ are the parameters to be estimated. The subscript i represents the cross-sectional units (i = 1,…, N), t indicates the time dimension (t = 1,…, T), and ε is the disturbance term.

Nevertheless, the presence of cross-sectional dependence can undermine the reliability of the system GMM estimator. To address this and ensure the robustness of the results, the study employs second-generation estimators. In particular, the robust standard errors approach developed by Driscoll and Kraay (1998) is advantageous, as it explicitly accounts for cross-sectional dependence. This estimator remains valid even in the presence of heteroskedasticity, autocorrelation, or interdependence across panels. Furthermore, the Driscoll–Kraay method has the added advantage of accommodating missing observations. Its applicability is especially relevant in contexts where the time dimension is relatively short compared to the cross-sectional dimension, as is the case in this study. Accordingly, Equations 3, 4 are estimated as a robustness check, incorporating fixed effects as shown below:

EFPi,t=λ+α0RECi,t+ α1FFCi,t+α2ECi,t+nXi,t+eit,(3)

GHGi,t=λ+α0RECi,t+ α1FFCi,t+α2ECi,t+nXi,t+eit.(4)

4 Results and discussion

4.1 Presentation of estimation results

The results of the dynamic panel estimation are reported in Table 3, with ecological footprint (EFP) and GHG emissions used as indicators of environmental sustainability. REC exhibits a negative relationship with environmental degradation in both models. In model 1, where environmental sustainability is proxied by ecological footprint, renewable energy has a statistically significant effect at the 5% level. In model 2, which uses GHG emissions as the dependent variable, the coefficient remains negative but is statistically insignificant. The estimated coefficients suggest that a one-unit increase in REC reduces the ecological footprint by approximately 0.00298 units and GHG emissions by about 0.00462 units. In contrast, non-renewable energy variables, including fossil fuel consumption and EC, are statistically insignificant across both models.

Table 3. The results of dynamic panel estimation using the system generalized method of moments.

The values in parentheses denote the robust standard errors; ^***p < 0.01, ^**p < 0.05, ^*p < 0.1. Source: Author computation.

The lagged dependent variables are positive and highly significant in both specifications, with coefficients of 0.784 (p < 0.01) in model 1 and 0.647 (p < 0.01) in model 2. These results indicate strong persistence effects, implying that past environmental conditions significantly influence current levels of ecological footprint and GHG emissions. Economic growth, proxied by GDP per capita, shows differing effects across the models. In model 2, GDP per capita has a positive and statistically significant coefficient (0.000489, p < 0.01), while in model 1 the coefficient is negative but statistically insignificant. Population growth also displays a positive and statistically significant effect in model 2 (0.681, p < 0.01), whereas its coefficient in model 1 is negative and statistically insignificant. Industrialization has positive but statistically insignificant coefficients in both models, with estimated effects of approximately 0.00120 in model 1 and 0.00400 in model 2. Regarding serial correlation diagnostics, the AR(1) test is statistically significant in model 1, which is expected due to first differencing in the system GMM framework. However, the AR(2) test is statistically insignificant in both models, indicating the absence of second-order serial correlation in the error terms. Furthermore, the Hansen J-test is statistically insignificant in both specifications, confirming that the instruments used are valid and appropriately specified.

The robustness check results are reported in Table 4. Overall, the findings show partial consistency with the main system GMM estimates. REC, economic growth, and population growth remain statistically significant, particularly in model 2. However, unlike the baseline results, fossil fuel consumption becomes statistically significant and exhibits a positive coefficient in the robustness estimations, indicating a worsening effect on environmental quality. Differences in statistical significance between the system GMM and Driscoll–Kraay estimations reflect the distinct estimation frameworks and underlying assumptions of the two methods. System GMM is retained as the primary estimator because it explicitly accounts for endogeneity, unobserved country-specific effects, and the dynamic structure of the panel data. In contrast, the Driscoll–Kraay approach is applied as a robustness check and provides standard errors that are robust to heteroskedasticity, serial correlation, and cross-sectional dependence. Consequently, some variables, including fossil fuel consumption, appear statistically significant under the Driscoll–Kraay specification but not under the more restrictive system GMM framework. Accordingly, the Driscoll–Kraay estimates are reported as supplementary evidence, while the system GMM results pinpoint the main basis for empirical inference.

Table 4. Result of the dynamic panel estimates using the Driscoll–Kraay specification.

The values in the parentheses denote the robust standard errors; ^***p < 0.01, ^**p < 0.05, ^*p < 0.1. Source: Author computation.

4.2 Discussion of estimation results

The results of our dynamic panel analysis provide strong evidence on how renewable and non-renewable energy consumption, economic growth, population growth, and industrialization affect environmental sustainability in SSA. Using ecological footprint (EFP) and GHG emissions as measures of environmental sustainability, we find that renewable energy consistently reduces environmental degradation. These findings are consistent with previous studies. For instance, Diallo and Ouoba (2024) report that countries such as Kenya and Ethiopia, which have higher shares of renewable energy, achieve improved environmental outcomes. Similarly, Abbasi et al. (2022) and Hou et al. (2023) show that renewable energy adoption has a clear negative effect on emissions, highlighting a consistent global pattern. On the other hand, non-renewable energy, particularly fossil fuel use, appears to increase environmental degradation in our robustness checks. This supports the findings of Hou et al. (2023) and Abbasi et al. (2022), emphasizing that reliance on fossil fuels continues to harm the environment and underscores the need for energy transition strategies. Interestingly, EC in SSA did not show a significant effect, possibly due to differences in energy mix, usage scale, and grid efficiency across countries.

Economic growth presents a mixed picture. In model 2, it is positively associated with environmental degradation though this is not significant in model 1. This significant positive nexus thus reflects the challenge of achieving “green growth” in SSA. This aligns with studies in China (Abbasi et al., 2022), OECD countries (Hou et al., 2023), and Turkey (Akadiri et al., 2019), which show that economic expansion often increases emissions in contexts reliant on fossil fuels. Population growth similarly raises environmental pressure, significantly increasing GHG emissions, though its effect on EFP is weak and statistically insignificant. This suggests that demographic pressures in SSA mainly influence environmental quality through energy demand and resource use rather than immediate ecological footprint measures. Industrialization shows positive but statistically insignificant effects on environmental degradation in both models, indicating a relatively limited impact in SSA compared to more industrialized regions. This contrasts with findings from more industrialized contexts such as Hou et al. (2023) and Ahmad et al. (2023), where industrial activity strongly affects emissions. The muted effect in SSA may be due to the prevalence of informal and small-scale industries that consume less energy. Meanwhile, the significant lagged dependent variables in both models highlight persistence effects, meaning that past environmental pressures strongly influence the current outcomes. This aligns with the understanding of cumulative environmental impacts observed globally, as noted by Achuo et al. (2022).

Comparing our findings with previous research, some clear patterns emerge. First, renewable energy consistently reduces environmental degradation across regions, including SSA (Diallo and Ouoba, 2024), OECD countries (Hou et al., 2023), China (Abbasi et al., 2022), India (Ahmad et al., 2023; Alola et al., 2023), and global studies (Achuo et al., 2022). Second, fossil fuel consumption and economic growth repeatedly appear as significant contributors to environmental pressure, illustrating a common challenge for sustainable development. Third, population growth and industrialization show context-specific effects, with a weaker environmental impact in SSA compared to more industrialized or densely populated regions, reflecting the region's unique structural and institutional characteristics. The findings therefore highlight the urgent need for SSA countries to expand renewable energy deployment through subsidies, regulatory support, and investment incentives. This will henceforth reduce ecological degradation while promoting inclusive and resilient growth. The consistency of these findings with prior studies underscores that renewable energy adoption is key to achieving SDG 7 and broader sustainability goals. At the same time, reducing reliance on fossil fuels is essential to limit long-term environmental pressures.

5 Conclusion and policy recommendations

This study examines the role of renewable and non-renewable energy, economic growth, population growth, and industrialization on environmental sustainability in SSA. The results show that REC plays a significant role in reducing environmental degradation, particularly when measured by EFP. In contrast, fossil fuel use and economic growth tend to increase environmental pressure, especially in GHG-based models, highlighting the difficulty of balancing development with ecological sustainability in the region. Population growth presents mixed effects, increasing environmental pressure when measured by GHG emissions but showing no significant impact on EFP. Industrialization, although positively linked to environmental degradation, does not appear to be a statistically significant driver in either model. The significant lagged effects further indicate that past environmental conditions continue to influence current sustainability outcomes. Overall, these findings provide strong evidence that expanding renewable energy adoption is critical for reducing ecological stress and mitigating GHG emissions in SSA.

The findings of this study emphasize the vital role of renewable energy in curbing environmental degradation in SSA. Policies should therefore focus on expanding renewable energy sources such as solar, wind, and small-scale hydropower, which have been shown to significantly lower both ecological footprint and GHG emissions. For example, our results indicate that a one-unit increase in REC reduces EFP by about 0.003 units and GHG emissions by 0.005 units, demonstrating the real impact of scaling up clean energy. Achieving this will require a mix of targeted subsidies, regulatory incentives, public–private partnerships, and programs that encourage community-level adoption alongside private sector investment.

At the same time, the positive effect of fossil fuel consumption on environmental degradation highlights the need to gradually reduce reliance on fossil fuels, particularly in power generation and transportation. Policymakers should promote energy efficiency, cleaner fuel alternatives, and ensure that the transition to greener energy does not limit access for vulnerable populations. Economic growth, while essential, can also increase environmental pressure, underlining the need for “green growth” strategies that separate development from ecological harm. Investments in energy-efficient infrastructure, clean technologies, and sustainable industrial practices are critical to achieving this balance.

Population growth further contributes to environmental pressure, especially when measured through GHG emissions, suggesting that policies targeting demographic factors such as family planning, education, and sustainable urban planning should be integrated with energy and environmental initiatives. Given the diversity across SSA countries, policies cannot be uniform; they must be adapted to local conditions, considering resource availability, institutional capacity, governance quality, and financial constraints. By aligning interventions with these empirical insights and regional realities, SSA can pursue sustainable energy transitions that reduce environmental degradation while supporting inclusive and resilient economic development.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/frevc.2026.1762584/full#supplementary-material

Supplementary material Supplementary material at the source

Footnotes

Edited by: Fulvio Fontini, University of Salento, Italy

Reviewed by: Jorge Cunha, University of Minho, Portugal

Mohamed Kadria, University of Sousse, Tunisia