Assessing the Environmental Sustainability Corridor in South Africa: The Role of Biomass Energy and Coal Energy

Abstract

South Africa’s national development plan has outlined aspirations to achieve a sustainable environment. However, the country remains bound for an unsustainable trajectory. Despite this ecological issue, no studies have probed how biomass and coal energy impact ecological quality. In light of this gap, this study inspects the environmental effect of political risk, coal energy, and biomass energy in South Africa. Also, this study integrates economic growth and natural resources into its framework. This study uses the load capacity factor (LC), which is a more aggregate proxy of ecological quality due to its accounting for the demand and supply aspect of the environment. This study uses the dynamic autoregressive distributive lag estimator (ARDL), which is capable of not only providing details of the influence of each determinant on LC in the long and short term but also of capturing the counterfactual shock of positive or negative exogenous variables on the LC. The kernel regularized least squares (KRLS) method is used for a robustness analysis of the dynamic ARDL approach. Furthermore, the findings of the dynamic ARDL simulation estimator disclose the negative impact of economic growth on the LC, thereby contributing to environmental deterioration by 0.552%. Natural resources and coal energy have an adverse impact on the LC, indicating a reduction in environmental sustainability by 0.037% and 0.290%, respectively. Meanwhile, biomass contributes to the LC, thereby promoting ecological quality by 0.421%. Political risk contributes to the reduction in the LC. This research provides pertinent policy considerations for policymakers and governments in South Africa, suggesting that the government of South Africa should invest in biomass energy and sustainable extraction procedures since biomass energy has a vital role in increasing ecological quality.

1. Introduction

Following the industrial revolution, the global surge in energy demand has intensified the exploitation of natural resources (NRs), driven by rapid population growth and economic development [1]. This heightened reliance on energy, particularly fossil fuels, has significantly exacerbated ecological degradation, as evidenced by numerous studies emphasizing the detrimental effects of energy consumption on environmental quality [2,3,4]. While energy consumption remains a critical catalyst for economic growth, its heavy dependence on nonrenewable sources, such as coal, oil, and gas, accelerates environmental deterioration. However, a substantial reduction in ecological deterioration can be achieved by shifting towards sustainable energy sources, thereby presenting a viable means to alleviate environmental harm [5,6]. Consequently, it is imperative to prioritize the pursuit of sustainable development, which can be realized through the formulation of environmental legalization. As a result, the United Nations Sustainable Development Goals (SDGs) were formulated with the goal of achieving these objectives. Sustainable development emerges as a global concern that assumes a pivotal role in safeguarding the existence of humanity [7,8]. In the contemporary milieu, the escalating utilization of fossil fuels poses a formidable impediment for nations striving to attain SDG 7, which centers on access to clean and affordable energy sources [9]. The nature and magnitude of energy production and consumption, essential for sustaining productive endeavors within countries, are widely recognized to exert a profound influence on the realization of ecological quality [8,9,10]. Concerning the challenges faced by most countries in meeting SDG 12, pertaining to responsible consumption and production, as well as SDG 13, addressing climate action, a primary driving factor underlying this predicament lies in the excessive depletion of resources during industrialization, which played a fundamental role in fostering nations’ economic growth, and which may not exert ecological deterioration [11,12]. However, the extraction process poses formidable threats, including deforestation, biodiversity loss, and the exacerbation of global warming [13,14,15].

Political stability is crucial for effective ecological regulation and mitigating environmental degradation, particularly in democratic nations where people have the freedom to advocate for higher ecological standards and hold political leaders accountable [9]. Additionally, institutional quality may significantly influence environmental excellence, with stronger political institutional quality associated with lower pollution levels and higher ecological quality [16], while ineffective institutional and governance management, including corruption, can hinder the enforcement of strict environmental regulations and contribute to illegal activities and environmental deterioration [17].

The above discussion draws the conclusion that developing effective sustainable environmental policies necessitates consideration of political stability, natural resource extraction, and energy consumption patterns. Moreover, limited to no studies have been conducted on any Sub-Saharan African nation. Furthermore, the determinants of the load capacity factor (LC), which is an aggregate proxy of ecological quality due to its accounting for the demand and supply aspect of the environment, have received less attention in Sub-Saharan African nations. As a result, as a major emerging nation situated in Sub-Saharan Africa, South Africa is the obvious choice for this analysis. In addition, the nation’s National Development Plan has outlined aspirations to achieve a sustainable environment, including a low-carbon economy. However, the country remains deeply rooted in an unsustainable trajectory. For instance, South Africa’s ecological deficit is relatively high given that the demand aspect of the environment (ecological footprint) is 3.21 gha per capita, while the supply aspect (biocapacity) is 1.28 gha per capita [18]. Notably, this is due to South Africa’s energy composition, which is heavily influenced by coal, which holds a dominant position in the overall energy supply. Coal energy dominates South Africa’s energy supply, accounting for 92% of electricity and 57% of heat generation, while renewables contribute only 6.6% to the total energy supply, with biomass making up the majority at 85%, being primarily used for heat production in residential and industrial sectors [19]. Furthermore, South Africa is blessed with abundant natural resources, which is manifest in its extractive economic framework, characterized by the significant underground extraction of coal, platinum, and gold. In addition, the country has been plagued with governance problems, corruption, and inefficient administration, which contribute to South Africa’s substantial political risk. Hence, having accounted for these issues, one could conclude that South Africa is an ideal candidate for this investigation.

Within this context, the objectives of this research are as follows. (i) This research ascertains the influence of economic expansion on the LC. (ii) This research inspects the influence of political risk on the LC. (iii) This research inspects the influence of natural resources on the LC in South Africa. (iv) This research ascertains the effect of coal and biomass energy on the LC. (v) This research determines the existence of a causal relationship between the LC and these variables (political risk, economic expansion, coal energy, natural resources, and biomass energy). The objectives of this study also highlight a major contribution of this study since we are unaware of any study that scrutinizes the environmental effect of political risk, economic expansion, coal energy, natural resources, and biomass energy within the framework of an emerging and Sub-Saharan African nation such as South Africa.

Another contribution of this study is the implementation of the dynamic ARDL stimulation method, which offers a comprehensive insight into both the short- and long-term impacts of key determinants, such as economic growth, political risk, and energy consumption, on the load capacity factor (LC), a nuanced proxy for ecological quality that accounts for both environmental demand and supply. This method enables the analysis to go beyond traditional techniques, allowing for a more precise evaluation of how positive or negative shocks to exogenous variables influence ecological outcomes over time. By capturing the interplay between these factors, this study provides a deeper understanding of the complex relationships underlying environmental quality in South Africa, a nation characterized by its unique mix of economic growth aspirations, energy-intensive industries, and environmental vulnerabilities. Additionally, the kernel regularized least squares (KRLS) method was used in this study as a robustness analysis for the dynamic ARDL approach.

Furthermore, this study’s findings hold significant implications for shaping South Africa’s energy policies and environmental strategies. This research underscores the urgent need for a diversified energy mix, emphasizing the expansion of renewable energy sources like solar and wind while reducing the dominance of coal in electricity and heat generation. Additionally, it highlights the importance of strengthening governance structures and reducing corruption to create a more stable political environment conducive to enforcing ecological regulations. These recommendations are not only critical for mitigating environmental degradation but also for aligning South Africa with the global sustainability agenda, including the achievement of SDG-7 (affordable and clean energy) and SDG-13 (climate action). Ultimately, this study’s contributions extend beyond South Africa, offering valuable insights for other emerging economies facing similar challenges and reinforcing the global discourse on sustainable development and ecological preservation.

2. Literature Review

Lately, researchers have investigated the growth–energy–pollution nexus from a broad to a detailed level. Also, incorporating the role of macroeconomic variables has been increasing in recent times. However, the outcome of the literature is contradicting. As a result, the subsection that follows provides an overview of the relationship between ecological sustainability and its determinants concerning the objective of this research.

2.1. Economic Growth (GDP), Coal Energy (CO), and Ecological Degradation

The energy–growth–pollution interaction in India for the timeline from 1970 to 2017 was investigated by [20]. They employed dynamic ARDL estimators, which confirmed that GDP increases the ecological footprint (EF), whereas CO induces the ecological footprint. For South Africa, ref. [21] considered the influence of GDP and CO on CO₂ emissions for the timeline between 1965 and 2017, employing the FMOLS method. They established that EKC is valid in South Africa and CO increases CO₂ emissions, respectively, within the period of study. In a study performed on developed and developing nations, ref. [22] reported a similar empirical result using the Augmented Anderson–Hsiao (AAH) two-step GMM estimator approach, in which GDP reduces CO₂ emissions in the entire sample and developing nations. However, in developed nations, GDP increases CO₂ emissions. Furthermore, their findings also revealed that CO₂ emissions increase as CO surges in both developed and developing nations. Utilizing the ARDL approach, ref. [23,24] confirmed that GDP and CO increase CO₂ emissions in South Africa. However, ref. [25] also probed into the effect of GDP and CO on CO₂ emissions for the period between 1966 and 2011 in India. They found that GDP and CO add to CO₂ emissions in India. Ref. [26] probed into the roles of GDP and CO on CO₂ emissions in China. They confirmed that GDP and CO promote CO₂ emissions. Ref. [27] inspected the effect of globalization on CO₂ emissions in seven selected nations. The authors established that a positive link between globalization and CO₂ emissions exists in those nations.

2.2. Biomass Energy (BM) and Ecological Degradation

In the study by [28], it was reported that biomass energy was negatively interconnected with CO₂ emissions in Germany based on an analysis of a dataset from 1965 to 2019 employing a machine learning algorithm. Utilizing the dynamic ARDL approach, ref. [29] probed into the relationship between BM and the EF in India for a dataset between 1970 and 2018; they uncovered that BM exerts an adverse effect on the EF. Meanwhile, ref. [30] assessed the effect of BM on the EF in G7 nations during the timeline from 1980 to 2016 employing the dynamic seemingly unrelated regression approach. They established a causal association between BM and the EF. The investigation by [31] in BRICS validated that the relationship between GDP and the EF is positive, but an adverse interconnection was suggested between the EF and BM over the period between 1992 and 2018. Ref. [32] considered the BM and load capacity factor interaction. The empirical outcome revealed that BM induces the load capacity factor (LC) in the USA. Also, in the research by [33] inspected the interaction between different forms of renewable energy and CO₂ emissions in China, and the empirical outcome indicated that BM reduces CO₂ emissions in China. Furthermore, the study by [34,35] probed into the relationship between BM and the EF in Malaysia and uncovered that BM exerts an adverse effect on the LC.

2.3. Natural Resources and Ecological Degradation

Due to the obvious significance of environmental issues, a majority of researchers have scrutinized the environmental effects of natural resources (NRs), which are recognized to have a major detrimental role in environmental quality. Numerous investigations have been performed to evaluate the environmental effect of NRs. For BRICS nations, ref. [36] considered the effect of NRs on CO₂, employing the MMQR, FMOLS, and DOLS approaches, which established that NRs increase CO₂ emissions. Ref. [37] inspected the linkage between NRs and the EF in 10 top emitting nations based on an analysis of a dataset from 2001 to 2021 employing the Driscoll and Kraay method and AMG, which uncovered a negative effect of NRs on the EF. However, by adopting a series of panel estimators to analyze the dataset of selected nations, ref. [38] scrutinized the effect of NRs on the LC. They uncovered a decreasing effect of NRs on the LC, thereby decreasing ecological quality. Utilizing the AMG and CCEMG method to assess a dataset of 35 South Asian economies collected between 1990 and 2020, ref. [39] evaluated the NR and EF nexus and reported that NRs exert a positive effect on the EF. For E7 nations, ref. [40] identified a decreasing effect of NRs on the EF during the period between 1993 and 2017 by employing the AMG method. However, ref. [41] assessed the NR and EF nexus and found that NRs exert a positive impact on the EF in seven selected nations. Moreover, ref. [42] appraised the relationships between NRs and the EF in BRICS and discovered a positive linkage between NRs and the EF. However, ref. [12] confirmed that NRs decrease CO₂ emissions in BRICS nations. A similar outcome was found in the study performed in Pakistan by [43].

Conversely, the work of [44] discovered that NRs contribute to ecological degradation in African nations by utilizing the FMOLS and DOLS methods. Likewise, ref. [45] inspected the interconnection between NRs and environmental degradation and demonstrated that NRs increased environmental degradation in BRICST economies. Ref. [46] probed into the effect of NRs on CO₂ emissions in China, adopting a dataset between 2000Q1 and 2020Q4. They confirmed that NRs increase CO₂ emissions. Ref. [7] uncovered the increasing impact of NRs on CO₂ emissions in developing nations.

2.4. Political Risk and Ecological Degradation

Over time, numerous researchers have inspected the environmental role of political risk (PR); nevertheless, their studies have produced conflicting results. The effect of PR on CO₂ emissions in the Netherlands was considered by [47]. Employing the Fourier approach, they confirmed that PR decreases CO₂ emissions. Ref. [48] uncovered that political stability reduces CO₂ emissions in BRICS economies. However, another study on BRICS by [17] confirmed that political stability increases the EF. Using four selected nations as a focus of the investigation, the study by [49] found that political stability decreases CO₂ emissions. Using MENA’s dataset, which spans the period between 2002 and 2018, ref. [50] identified an insignificant link between CO₂ emissions and political stability in the lower and middle quantiles, but in the upper quantiles, political stability decreased CO₂ emissions. The empirical outcome of the study by [51] revealed that PR induced ecological degradation in South Asian economies during the period from 1996 to 2019. Also, ref. [52] scrutinized the interaction between PR and the LC, and the empirical outcome revealed that PR induces the LC in Brazil. Ref. [53] studied the effect of PR on CO₂ emissions in Canada. The authors established a negative interaction between PR and CO₂ emissions.

After appraising several studies, we also discovered that the effects of CO energy, natural resources, political risk, and biomass energy on the environment are under-researched, particularly for specific Sub-Saharan African nations. Furthermore, no prior research has investigated the role of CO energy, natural resources, political risk, and biomass energy on the load capacity factor for the case of South Africa, presenting a significant void in the literature. As a result, we attempted to propose major policy guidance in this research depending on the aims stated above.

3. Data and Model Specification

3.1. Data

This study examined the role of economic growth, CO energy, biomass energy, natural resources, and political risk on the load capacity factor, in which South Africa is the focus. This study applied an annual dataset spanning between 1984 and 2022. The starting period of 1984 was selected due to data availability for political risk. The load capacity was obtained from the database of the global footprint network, whereas the datasets of natural resources and economic growth were obtained from the World Bank’s database. Biomass energy was sourced from the material flow database, and, finally, the dataset of CO energy was extracted from the British Petroleum statistics database. This research converted all series into their natural logarithms to improve the analysis’s reliability and normalize the data. This enabled the coefficients to be understood as elasticities. The parameters and descriptions are presented in

Table 1. Description and sources of the variables.

Indicators	Description of the Variable Used	Metric	Sourced
LC	Load capacity factor	Hectares per capita	Global Footprint
GDP	Economic growth	Constant 2015 USD	World Bank Database
CO	CO energy	Exajoules	British Petroleum database
BM	Biomass energy	Tonnes	Material Flows Database
NRs	Natural resources	% of GDP	World Bank Database
PR	Political risk	Index	PRS Database

.

Based on the discussion above, economic growth, coal energy, natural resources, political risk, and biomass energy can have an adverse or positive impact on environmental sustainability. The load capacity factor was used as a metric for environmental sustainability in this study, while economic growth, CO energy, natural resources, political risk, and biomass energy were the independent variables. In line with prior studies such as [34,54,55,56,57], this study developed the economic function, which is specified as follows:

$${LCAP}_t={\vartheta }_0+{\vartheta }_1{GDP}_t+{\vartheta }_{2}CO+{\vartheta }_3{BM}_t+{\vartheta }_4{NR}_t+{\vartheta }_5{PR}_t+{\epsilon }_t$$

(1)

where LC, CO, BM, GDP, NRs, and PR depict the load capacity factor, CO energy, biomass energy, economic growth, natural resources, and political risk. The subscript t symbolizes the period of study (1984–2022), and the error term is indicated as $\epsilon$.

We anticipated a decreasing effect of GDP on the LC, i.e., $\left({\vartheta }_1={\displaystyle \frac{\delta LCAP}{\delta GDP}}<0\right)$. Previous studies such as refs. [58,59] agree with this perspective. We anticipated a negative connection between CO and the LC, i.e., $\left({\vartheta }_2={\displaystyle \frac{\delta LCAP}{\delta CO}}<0\right)$, which is echoed by previous studies [26,29]. Centered on prior studies [31,32], this research anticipated an increasing effect of BM on ecological sustainability, i.e., $\left({\vartheta }_3={\displaystyle \frac{\delta LCAP}{\delta BM}}>0\right)$. Concerning the effect of natural resources on the LC, we anticipated an inverse effect. This perspective was developed based on the studies by [60,61], i.e., $\left({\vartheta }_4={\displaystyle \frac{\delta LCAP}{\delta NR}}<0\right)$. Based on the research by [62], this research anticipated a decreasing effect of political risk on ecological sustainability, i.e., $\left({\vartheta }_5={\displaystyle \frac{\delta LCAP}{\delta PR}}<0\right)$.

3.2. Econometric Strategy

The conventional (PP and ADF) unit root is applied in this study to detect the stationary nature of the parameters used. Nonetheless, having received widespread criticism regarding the outcomes of conventional stationary techniques, we chose to employ the Zivot and Andrews unit root test developed by [63], since it is recognized for delivering accurate outcomes and accounting for at least one break. It is critical to remember that this approach is only employed to determine whether the series have unit root issues at various levels, while other tests cannot discriminate between structural break and unit root issues.

Furthermore, this study used the dynamic ARDL estimator to establish the long-run association. The ARDL necessitates the determination of an appropriate lag, and the potential issue of endogeneity was addressed by selecting an optimal lag period. The estimator is applicable when the series are 1(0) or 1(1) or both 1(0) and 1(1). The selection of an appropriate lag helps fix the issue of possible multicollinearity in the estimation [64,65]. Also, the ARDL can produce the effect of the LC’s determinants in the long term and in the short term. Moreover, the ECT displays the convergence information. As defined in the study by [64], the ARDL equation in relation to this study’s variables is shown below:

$$\begin{array}{cc}\hfill \mathit{\Delta }{LCAP}_t={\theta }_0+& {\displaystyle \sum _{l=1}^p}{\theta }_1\mathit{\Delta }{LCAP}_{t-1}+{\displaystyle \sum _{i=1}^p}{\theta }_2\mathit{\Delta }{GDP}_{t-1}+{\displaystyle \sum _{i=1}^p}{\theta }_3{COAL}_{t-1}+{\displaystyle \sum _{i=1}^p}{\theta }_4{BIO}_{t-1}+{\displaystyle \sum _{i=1}^p}{\theta }_5\mathit{\Delta }{NR}_{t-1}\hfill \\ & +{\displaystyle \sum _{i=1}^p}{\theta }_6\mathit{\Delta }{PR}_{t-1}+{\pi }_1{LCAP}_{t-1}+{\pi }_2{GDP}_{t-1}+{\pi }_3{COAL}_{t-1}+{\pi }_4{BIO}_{t-1}+{\pi }_5{NR}_{t-1}+{\pi }_6{PR}_{t-1}\hfill \\ & +{ϵ}_t\hfill \end{array}$$

(2)

In Equation (3), ${\theta }_i(i=1\dots 6)$ denotes the coefficient of the parameter in the short term and ${\pi }_i (i=1\dots 6)$ demonstrates the long-term relationship among the parameters. t denotes lag lengths. Equation (3) is modified into Equation (4) by including the ECM in the ARDL, which is defined in the work by [66] as follows:

$$\begin{array}{cc}\hfill \mathit{\Delta }{LCAP}_t={\theta }_0+& {\displaystyle \sum _{l=1}^p}{\theta }_1\mathit{\Delta }{LCAP}_{t-1}+{\displaystyle \sum _{i=1}^p}{\theta }_2\mathit{\Delta }{GDP}_{t-1}+{\displaystyle \sum _{i=1}^p}{\theta }_3{COAL}_{t-1}+{\displaystyle \sum _{i=1}^p}{\theta }_4{BIO}_{t-1}+{\displaystyle \sum _{i=1}^p}{\theta }_5\mathit{\Delta }{NR}_{t-1}\hfill \\ & +{\displaystyle \sum _{i=1}^p}{\theta }_6\mathit{\Delta }{PR}_{t-1}+{\pi }_1{LCAP}_{t-1}+{\pi }_2{GDP}_{t-1}+{\pi }_3{COAL}_{t-1}+{\pi }_4{BIO}_{t-1}+{\pi }_5{NR}_{t-1}+{\pi }_6{PR}_{t-1}\hfill \\ & +{\pi }_7\alpha {ECT}_{t-1}+{ϵ}_t\hfill \end{array}$$

(3)

where the error correction term is given by ${ECT}_t$, indicating the adjustment rate of balance in the long term from a shock in the short term, and the coefficient of ${ECT}_t$ is expected to be negatively significant. The ARDL approach was applied to scrutinize the linkage between CO₂ emissions and its regressors, after ascertaining the cointegrating interconnection in Equation (4). Furthermore, the findings of the dynamic ARDL stimulation method were verified using the kernel-based regularized least squares (KRLS) method.

4. Findings and Discussions

For the pre-estimation assessment in this research, we try to understand the basic nature (descriptive statistics) of the observed parameters, which is summarized in

Table 2. Descriptive statistics.

	LC	GDP	CO	BIO	NRs	PR
Mean	−0.378	3.722	0.510	8.161	0.703	1.501
Median	−0.376	3.718	0.520	8.164	0.679	1.520
Maximum	−0.298	3.796	0.594	8.236	1.078	1.580
Minimum	−0.465	3.630	0.405	8.018	0.395	1.295
Std. Dev.	0.053	0.058	0.057	0.044	0.149	0.063
Kurtosis	−0.086	−0.063	−0.191	−0.942	0.350	−1.306
Skewness	1.500	1.434	1.702	4.158	3.121	4.834

. BIO (8.161) has the highest average value next to GDP (3.722), PR (1.501), NRs (0.703), and CO (0.510), while the LC has the lowest mean value of −0.378. The range of observed variables are as follows: LC (−0.465 to −0.298), GDP (3.630 to 3.630), CO (0.405 to 0.594), BIO (8.018 to 8.236), NRs (0.395 to 1.078), and PR (1.295 to 1.580). The standard deviations of the LC, GDP, CO, BIO, NRs, and PR are 0.053, 0.058, 0.057, 0.044, 0.149, and 0.063, respectively. For skewness, the outcome shows that all series are positively skewed. Moreover, for kurtosis, we uncovered that LC, GDP, and CO are platykurtic in nature, while BIO, NRs, and PR are leptokurtic in nature.

Furthermore, in this study, we utilized a battery of conventional stationary tests (PP and ADF) to uncover the integration order of the series, in which the findings are indicated in

Table 3. Conventional unit root.

	ADF		PP
	I(0)	I(1)	I(0)	I(1)
LCF	−0.134	−7.702 *	−2.697	−9.412 *
GDP	−2.475	−3.894 **	−2.424	−3.850 **
CO	−2.114	−7.817 *	−2.114	−9.872 *
BM	−6.429 *	−8.094 *	−6.429 *	−8.720 *
NRs	−3.100	−8.669 *	−3.100	−8.669 *
PR	−2.643	−5.350 *	−2.866	−5.345 *

Note: p < 1% and p < 5%, denote * and **.

. The findings disclose that biomass energy is stationary at levels. Moreover, all series are stationary at I(1). In a scenario when the series has a break, the normal unit root test (ADF and PP) will generate erroneous estimates. Time series data are exposed to instabilities due to macroeconomic and structural phenomena, like policies that may impede the variables’ stability, which pinpoint a specific economic occurrence. These issues could potentially have an impact on the outcomes of any investigation undertaken for any nation. Conventional unit root testing typically fails to take into account such disturbances during estimation. Thus, we used the Zivot and Andrew test, which was devised by [63].

Table 4. ZA unit root.

	I(0)	Break	I(1)	Break
LC	−2.032	2012	−9.986 *	2003
GDP	−2.466	2016	−5.687 *	2008
CO	−6.600 *	2008	−8.386 *	2010
BM	−7.322 *	1996	−9.211 *	1993
NRs	−4.584	2004	−9.811 *	2009
PR	−5.359 **	2009	−8.349 *	2009

Note: p < 1% and p < 5%, denote * and **.

summarizes the findings, which revealed that no unit root issues exist at any level for biomass energy, coal energy, and political risk. In addition, all series also are stationary at I(1). The results of the conventional stationary tests and the Zivot and Andrew test clearly indicate that there is a mixed order of integration among the variables used.

After evaluating the pattern of integration of the observed variables, we continued to assess the long-term interrelationship. We utilized the ARDL bounds test to detect the long-run interconnection. The findings of the ARDL bounds test are shown in

Table 5. Bounds ARDL cointegration.

Models	F-Statistic		T-Statistic
	16.089 *		−10.612 *
	Crit. Value (F-Statistic)		Crit. Value (T-Statistic)
Significance level	1(0)	1(1)	1(0)	1(1)
10%	2.26	3.35	−2.57	−3.86
5%	2.62	3.79	−2.86	−4.19
2.5%	3.41	4.68	−3.43	−4.79
1%	3.15	4.43	−3.43	−4.99

Note: * portrays p < 0.01.

, which disclose an indication of a long-run interrelationship between the LC and its independent variables (GDP, CO, BM, NRs, and PR) because both the T- and F-statistics are significantly higher than the critical values of Kripfganz and Schneider (LB and UB).

Upon the establishment of a cointegrating relationship between CO₂ and its determinants, this study moved on from there to assess the short-term and long-term relationships. The findings of the dynamic ARDL stimulation method are reported in

Table 6. ARDL.

	Long-Term			Short-Run
Variables	Coeffics	T-Statistic	Prob Value	Coeffics	T-Statistic	Prob Value
GDP	−0.552 *	−4.08	0.000	−0.890 *	−3.26	0.003
BM	0.421 *	2.90	0.008	0.401 *	5.44	0.000
CO	−0.290 ***	−1.82	0.080	−0.334 **	−2.43	0.022
NRs	−0.037 **	−2.08	0.048	0.008	0.41	0.687
PR	−0.226 **	−2.67	0.013	−0.224 **	−2.56	0.017
C	−1.211	−1.37	0.183	-	-	-
ECT (−1)	-	-	-	−0.899 *	−4.72	0.000
	Diagnostic Tests
J-B Normality	0.929 (0.628)
χ2 LM	1.446 (0.252)
χ2 RESET	1.038 (0.307)
χ2 BPG	0.703 (0.669)

Note: ***, **, and * represent p < 0.10, p < 0.05, and p < 0.01.

. GDP exhibits a negative and significant effect on the LC, indicating that an upsurge in GDP decreases the LC, thereby increasing ecological deterioration in the long and short terms. Thus, our result shows that a 1% surge in GDP reduces the LC by 0.552% (long run) and 0.890% (short run). Moreover, this association between the LC and economic expansion specifies that South Africa is currently in the scale phase, which is experienced by emerging nations when fossil fuel energy sources underpin economic expansion. This suggests that South Africa prioritizes economic expansion over environmental degradation. Several emerging nations have pursued this economic growth trajectory. Hence, this research’s outcomes will have significant policy implications for South Africa and other emerging economies. Thus, attaining the aspirations of SDG 13 in South Africa will indeed be challenging. This is an urgent call for stakeholders and policymakers to adopt a more deliberate stance on disassociating economic expansion from pollutant emissions in the nation’s energy basket. Our finding conforms with the research by [58,61] in BRICS economies, ref. [59] in Brazil, [67] in India, and [68] in OECD nations. Conversely, the research by [69] contrasts this outcome.

In addition, the effect of CO decreases the LC, considering that the coefficient of CO is 0.290 in the long term, while for the short term, is 0.334. This result is predictable given that coal energy exerts the most carbon-intensive fossil fuel during usage, emitting 72% more climate-changing CO₂ per unit of energy than gas. Despite its target of achieving ecological sustainability, South Africa continues to consume more fossil fuel energy sources, most especially CO, to meet the increasing energy demand. Coal has been the principal energy source for South Africa’s manufacturing-based economy for the past century. Over 90 percent of the electricity in South Africa is currently generated by coal because it is mined in South Africa. This result conforms with the research by [26] in China, ref. [70] in France, ref. [71] in G20 nations, and ref. [29] and ref. [20] in India.

Furthermore, our result shows that a 1% surge in BM induces the LC by 0.421% (long run) and 0.401% (short run) at the 1% significant level. Thus, biomass energy exhibits a positive interconnection with the LC in the long and short terms. This outcome indicates that biomass energy helps to achieve a sustainable environmental condition. This outcome agrees with prior studies [29,31,72,73,74,75]. This conclusion suggests that while it is unfeasible to eliminate the usage of fossil fuels, notably, coal, which seems to be the prevailing source of energy and a significant component of South Africa’s economic growth, the government can enhance environmental quality by increasing its utilization of biomass energy. Thus, the use of biomass energy could support the accomplishment of sustainable development. Therefore, policymakers and stakeholders need to restructure their energy basket by incorporating more greener energy. Furthermore, other prior studies, such as [28] in Germany, ref. [33] in China, and ref. [32] in the USA, also give credence to the positive role of BM on ecological quality. Meanwhile, the study by [30] offers a contradicting outcome by confirming the detrimental effect of BM on ecological quality in G7 nations.

This study’s findings show that natural resources exacerbate the degradation of the environment in South Africa. Based on these findings, when all other variables are held constant, an increase in NRs by 1% will lead to a reduction in the LC by 0.037% in the long term, indicating that the practices of natural resource extraction in South Africa have not been sustainable. It is not surprising to observe this outcome, as the pursuit of economic development in South Africa propels the nation towards industrialization and fosters the excessive exploitation of its natural resources. Consequently, this process depletes the availability of natural resources and accelerates ecological degradation, as overdependence on these resources hampers their regeneration, which diminishes biocapacity. Furthermore, the extensive coal mining in South Africa engenders profound ecological degradation within delicate natural ecosystems, including forests, and possesses the capacity to inflict irreparable scars upon the landscape, thereby resulting in detrimental consequences for humans, animals, and plants through the destruction of habitats and the contamination of the environment. This result conforms with the investigations by [76,77,78,79], positing that the escalating global demand for natural resources, driven by population expansion and economic progress, frequently outpaces the earth’s capability to regenerate them. As a result, resources are exploited at an unsustainable rate, causing depletion and environmental damage.

Our results reveal that a 1% upsurge in PR mitigates the LC by 0.226% (long run) and 0.224% (short run). Thus, this outcome highlights the adverse effect of PR on the LC in South Africa. This adverse association suggests that the current political climate hinders ecological sustainability in South Africa. Political instability engenders a circumstance wherein the integrity of institutions and the efficacy of governmental administration are compromised, thereby precipitating environmental degradation. Notably, corruption serves as a tangible exemplification of the diminished quality of institutions, as it has the potential to undermine the stringency of environmental regulations. This, in turn, fosters a rise in illegal manufacturing and extraction activities, thereby contributing to ecological deterioration. In addition, this result conforms with the investigations by [17,51], which reported the detrimental effect of PR on ecological quality in South Asian nations and BRICS nations, respectively. However, the positive role of PR on ecological quality was observed in the studies by [52] in Brazil, ref. [53] in Canada, and [47] in the Netherlands.

Furthermore, in Table 6, the results of the post-estimation analysis are disclosed, which suggest that the residuals of the model exhibit a normal distribution and do not suffer from serial correlation, misspecification, or heteroscedasticity issues. Figure 1 shows that the residual of this model is stable at a 5% level of significance.

The dynamic ARDL stimulation method has the ability to automatically generate insightful impulse–response plots, which reveal the counterfactual shock of 1% positive or negative change in the exogenous variables on the LC. The outcome of the dynamic ARDL stimulation method is presented in Figure 2, Figure 3, Figure 4, Figure 5 and Figure 6. Figure 2 shows the responses of the LC to a change in GDP. For the positive shock in GDP (see Figure 2A), a 1% positive change in GDP will lead to a decrease in the LC. Meanwhile, a 1% negative shock in GDP will induce the LC (see Figure 2B). Furthermore, the responses of the LC to ±1% changes in biomass energy are presented in Figure 3. A 1% surge in BM will lead to a rise in the LC (see Figure 3A), while a 1% decrease in BM will cause a decrease in the LC (see Figure 3B). In addition, Figure 4 shows the responses of the LC to ±1% changes in natural resources. Figure 4A shows that the LC declines due to a 1% rise in the LC. Likewise, Figure 4B shows that a 1% decline in NRs will cause a reduction in NRs. Furthermore, Figure 5 reveals the positive and negative changes in NRs on the LC. Figure 5A indicates that a 1% positive change in NRs reduces the LC. Conversely, Figure 5B depicts that a 1% decline in NRs decreases the LC. The responses of the LC to ±1% changes in political risk are showcased in Figure 6. A 1% positive shock in PR leads to an adverse impact on the LC, as highlighted in Figure 6A, while Figure 6B shows that a 1% negative change in PR mitigates the LC.

For the robustness analysis of the result of the dynamic ARDL method, this study conducted the kernel regularized least squares (KRLS) method.

Table 7. Result of the KRLS methods.

	Average	SE	t-Value	p-Value	P25	P50	P7
CO	−0.288 *	0.481	−5.987	0.000	−0.392	−0.298	−0.145
BM	0.241 *	0.053	4.555	0.000	0.053	0.155	0.447
GDP	−0.261 *	0.038	−6.758	0.000	−0.429	−0.258	−0.098
NRs	−0.007	0.012	−0.549	0.586	−0.039	−0.007	0.025
PR	−0.112 *	0.035	−3.162	0.003	−0.195	−0.119	−0.007

Note: Significance at the 1% is indicated with * symbols.

shows the results of the KRLS method. It shows that CO has an adverse effect on the LC, wherein a 1% surge in CO will lead to an average decline in the LC by 0.288%. Likewise, a 1% surge in GDP will cause an average drop in the LC by 0.261%. Thus, GDP decreases the LC in South Africa. However, natural resources have a negative and insignificant effect on the LC. Furthermore, PR causes a decline in the LC, wherein a 1% rise in PR will lead to an average decline in the LC by 0.112. Conversely, a 1% surge in BM will cause a rise in the LC by 0.241%. Thus, the KRLS method shows that CO energy, GDP, natural resources, and political risk exert an adverse influence on the LC, while biomass energy increases the LC in South Africa. The outcome from the KRLS method provides support for the result of the dynamic ARDL stimulation method.

5. Conclusions and Policy Implication

Identifying the effect of economic growth, coal energy, biomass energy, natural resources, and political risk on the load capacity factor is the major objective of this study. This study focused on South Africa, in which the dataset covers the timeline between 1984 and 2022. For the stationarity nature of the observed parameters, the mix order of integration was confirmed by the conventional unit root tests (ADF and PP) and the ZA unit root. Moreover, this study also used the Bound test approach, along with [80]’s critical approximation, which confirmed a long-term equilibrium interconnection between the LC and its regressors. Furthermore, the findings of the dynamic ARDL simulation estimator disclosed that a 1% surge in GDP reduces the LC by 0.552% (long run) and 0.890% (short run), thereby suggesting that economic growth contributes to environmental deterioration. The results further reported a decline in the LC by 0.290 in the long run due to a unit increase in CO. The result of natural resources has an adverse impact on the LCAP, indicating a reduction in environmental sustainability by 0.037%. Meanwhile, biomass contributes to the LC, thereby promoting ecological quality by 0.421%. Political risk contributes to a reduction in the LC by 0.226% (long run) and 0.224% (short run). The KRLS method also corroborated the impact of each regressor on the LC, except for natural resources.

The results of this study offer credence for South African authorities to increase R&D spending on biomass technologies. South Africa’s commitment to achieving ecological sustainability and producing a sustainable ecosystem by expanding the quantity of biomass energy in its energy market may be mainly ascribed to this milestone. Furthermore, firms specializing in producing and assembling biomass energy sources need to be eligible for incentives such as subsidies, price control regulations, and tax reliefs. This will empower these firms to increase their output level, thereby making products more affordable and available to all citizens. Furthermore, natural resources undermine environmental quality; therefore, the implementation of sustainable strategies for the effective assessment and oversight of resources is of critical importance for South Africa. Prioritizing a sustainable extraction and production process is a pivotal strategy, thereby positioning South Africa to effectively promote biodiversity, resource recycling, and enhance the overall ecological state while mitigating the depletion of biocapacity.

The establishment of a more robust institutional framework must be implemented by the South African authorities with the aim of mitigating the bureaucracy that hampers the efficient implementation of ecological policies. The institutional framework must be meticulously designed to facilitate the effective implementation and administration of ecological regulations. South African authorities can address the obstacles undermining the effective implementation of ecological regulations by cultivating robust cross-sectoral collaboration, optimizing administrative mechanisms, and reinforcing governance structures to ensure cohesive and efficient regulatory enforcement. Empowering regulatory authorities with the essential knowledge and expertise to proficiently address complex ecological challenges represents a strategic investment that ensures the successful realization of their objectives. This approach significantly mitigates risks associated with ineffective implementation, thereby facilitating the seamless execution of ecological regulations.

Prior to drawing any conclusions from this study, it is crucial to identify that no policy structure can encapsulate all appropriate policy issues within the borders of a specific framework, and even this investigation is not different. The model was designed by considering only five variables, which is this study’s drawback. Subsequent research should replicate similar investigations in other emerging and developing countries by examining the asymmetric influence of these variables or incorporating other socioeconomic indicators. Another limitation of this study is the coverage of the data adopted by this study. As a result, future studies can extend the period of study. Furthermore, another limitation of this research is that it relied on time series analysis using the ARDL estimator. However, future studies could adopt other time series approaches such as the quantile-on-quantile approach, wavelet tools, Fourier ARDL, and quantile ARDL.