Introduction

Among the biggest difficulties the global economy is currently experiencing are the problems of climate change and ecological deterioration. Threats to human health as well as to productivity and income levels are created by this problem1,2. Energy demand increases along with the expansion of economic activities, which increases greenhouse gas emissions that are detrimental to the ecosystem3. Practically every country has joined the Paris Agreement and decided to cut GHGs emissions as a way to address the worldwide climate change challenge. The countries’ dedication to accomplishing the aim is crucial to the mission's viability and the accomplishment of the sustainable development goals. Nations are encouraged to pursue green growth in order to cut the trajectory of CO2 emissions. As per4, the global economy would lose over 18% of its GDP due to climate change.

Regrettably, as stated in the UN Production Gap Report 2021, the present production plan exceeds the cap imposed by the Paris Agreement (United Nations Environment Program, 2021). Thus, all countries must take the necessary steps to guarantee that CO2 emissions can be drastically decreased and contribute to achieving sustainable development goals5,6,7. Although many countries have begun to turn to sustainable energy and renewable energy as alternatives to fossil fuels, the quantity needs to be improved to meet the growing energy demand. On the other hand, environmental-related technologies may be a panacea for the environmental issue since it aids in combining sustainable economic growth with improved ecological management8,9. This results from positive spillover effects from innovation efforts that lead to the development of new goods, processes, and techniques that can mitigate adverse environmental effects10,11. According to12, industrialized nations attained higher ecological quality due to cutting-edge technologies that greatly decreased emissions and enhanced ecological conditions.

Even with the expanding body of studies examining the variables that influence environmental quality, the majority of these studies have focused on industrialized nations because most innovation occurs in these nations13,14. Nonetheless, given develo** nations would be the most impacted by climate change, it is vital for policy measures to comprehend how environmental technologies impact CO2 emissions 15. The eco-friendly technologies needed to adapt to climate change would differ depending on local situations12. Thus, research on emerging nations is crucial for formulating policies supporting environmental technologies.

Considering the efficiency of these variables in achieving carbon neutrality targets, the primary purpose of our research is to concentrate on CO2 drivers by evaluating the impact of renewable energy, financial development, and environmental technologies between 1980 and 2020. South Africa is selected for the study due to many reasons. First, develo** nations such as South Africa have been overlooked in multiple ecological studies, and few research is accessible. The South African energy system’s cornerstone is coal, which is based on fossil fuels, providing approximately 70% of the installed power production capacity16. Nonetheless, the 2019 Integrated Resource Plan outlines a long-term diversification of the power mix by 2030 and strives to reduce the energy sector's carbon impact while supplying more energy and guaranteeing a socioeconomically equitable16. The government will have to make difficult decisions as it pursues its goals of diversifying and minimizing the detrimental ecological effect of the nation's energy mix. To do this, it must actively include the public in the discussion. Yet, South Africa's integrated policies, robust regulation, well-designed incentives for low carbon investment, particularly private investment, regional connectivity, and increased efficiency provide it with unmatched capability for the challenge. Based on the interesting facts regarding South Africa, the study formulates the research objectives by asking the following basic questions:

  1. 1.

    Can renewable energy consumption contribute to ecological sustainability in South Africa?

  2. 2.

    Does environmental-related technologies improve ecological sustainability in South Africa?

  3. 3.

    Does financial development improve ecological sustainability in South Africa?

The current study makes three major contributions. First, the present investigation will make a substantial and vital contribution to South Africa's policymakers in light of how climate change and biodiversity affect the country. Second, notwithstanding the expanding literature on technological innovation, the significance of environmental-related innovation has received less attention. Third, the study contributes methodologically by utilizing time-varying causality along with the autoregressive distributed lag model. The implementation of the novel econometric approach such as time-varying causality presents another dimension into these connections. Unlike the conventional causality tests such as17,18 causality tests, this technique, which uses a bootstrap** strategy to account for small-sample bias, enables us to identify any structural changes in the framework as well as the progression of causal interrelationships between sub-periods.

The following sections serve as the architecture and design for the present investigation: The assessment of the empirical studies is presented in Sect. 2. Section 3 provides the data and empirical strategies. Section 4 presents the results of the empirical research and their discussion. Section 5 outlines noteworthy policies based on the findings.

Theoretical framework and literature review

Awareness of the devastating effect of global warming and climate change is no longer new to develo** and developed economies. As a result, it is vital to find a long-term solution to this problem. This is evident in the discussions in Paris Agreement, COP26 and COP27, respectively. Nations have come together to find a solution by limiting their level of emissions. Several nations including South Africa have reaffirmed their commitment towards carbon neutrality. However, based on recent development, it looks unlikely that most of these nations will be able to attain their carbon neutrality targets. Several economic variables have been highlighted as the drivers of global warming and climate change. However, proposing a realistic and concrete policy regarding these economic variables' effect on environmental deterioration has faced several roadblocks. Studies have reported dissimilar findings regarding the drivers of ecological decline. Nonetheless, inconclusive results have surfaced which are attributed to the economic conditions of the nation/nations of investigation, techniques employed, and period of study.

With the motive of formulating SDGs policies for Finland19, evaluated the driving factors behind ecological deterioration using yearly data between 1990 and 2019. The study applied the novel Fourier approaches in order to propose necessary policies. The study results disclosed that environmental-related technologies and green energy play a vital role towards Finland's carbon neutrality target, while economic growth in Finland is not sustainable. Similarly14, used the Fourier approaches in order to evaluate the role of environmental-related technologies and green energy towards the USA carbon neutrality target within the EKC coffin. The investigators used yearly data from 1990 to 2019 to explore the nexus, and the study findings disclosed that ecological-related technologies and green energy lessen CO2 emissions while economic progress and financial development deter mitigation of CO2 emissions in the United States. Likewise13, in their investigation on the drivers of GHG emissions in the Nordic economies, consider the role of energy intensity, eco-innovations, raw material productivity and economic expansion. The investigators assessed the nonlinearity characteristics of the variables and, as a result, employed the panel nonlinear ARDL. The study results disclosed that positive (negative) shifts in energy intensity, eco-innovations and raw material productivity decrease (increase) GHG emissions, while positive (negative) in economic progress increase (decrease) GHG emissions.

The study of20 explores the CO2 emissions drivers for the 20 biggest economies within the EKC framework. The authors used data from 1990 to 2018 and economic growth, green energy and eco-innovations are considered drivers of CO2 emissions. The investigators used MMQR to explore this interrelation and the study findings disclosed that a decrease in CO2 emissions is attributed to the intensification of green energy and eco-innovations. In contrast, the intensification of economic expansion and fossil fuel causes an increase in CO2. With the aim of devising meticulous policies for the BRICS economies21, explore the effect of disintegrated energy, globalization and innovation on ecological footprint using data from 1990 to 2018. The studies used ample techniques-AMG and CCEMG estimators to evaluate this nexus. The study result unearths that mitigation of EF is caused by the intensification of renewable energy, globalization and innovation, while fossil fuel and economic expansion lessen ecological quality.

Likewise, using the Malysia case22, used the wavelet and FMOLS and ARDL estimators to analyze the nexus between CO2 and financial development, green energy, nonrenewable energy and economic expansion. The findings uncovered that dirty energy and economic expansion intensify CO2 emissions while green energy lessens CO2 emissions. Besides, an insignificant linkage exists between financial development and emissions, suggesting that Malaysia financial section is still young. Moreover, the study of23 for the MINT economies using data between 1990 and 2019 reported that an upsurge in financial development and economic expansion lessens the sustainability of the ecosystem while renewable energy and FDI inflows lessen ecological dilapidation. Furthermore, several studies in energy and environmental literature have reported significant various drivers of ecological quality such as economic growth, financial development, renewable energy, innovation, trade and globalization24,25,26. Likewise, several time series and panel studies also documented substantial studies on the nexus between economic growth, innovation, financial development and CO2 emissions27,28,29,30,31,32,33.

The current investigation explores the determinants of CO2 emissions, such as financial development, renewable energy, and environmental related technologies. Drivers of CO2 emissions have been investigated in previous research; nevertheless, these studies have certain distinctive gaps or shortcomings. With some appropriate modifications, the present investigation fills these gaps in the previous literature. First, previous researchers have developed a variety of viewpoints on the relationship between CO2 emissions, renewable energy, financial development, and environmental technologies. Conversely, the current investigation resolves the literature debate about the correlation between CO2 emissions, renewable energy, environmental technologies, and financial progress. The ecological implications of CO2 emissions via renewable energy, ecological technologies, and financial development have been studied in a variety of nations, but little is documented about the situation of South Africa. Assessing renewable energy, financial development, environmental technology, and their relationships with CO2 emissions in South Africa closes the gap in the literature. Secondly, many techniques and analytical approaches such as ARDL, fully modified OLS (FMOLS), dynamic OLS (DOLS), vector autoregressive (VAR), nonlinear ARDL and dynamic ARDL have been used in the past to investigate the drivers of CO2 emissions. Implementing the novel econometric approach, such as time-varying causality, presents another dimension to these connections. Unlike the conventional causality tests such as17,18 causality tests, this technique, which uses a bootstrap** strategy to account for small-sample bias, enables us to identify any structural changes in the framework as well as the progression of causal interrelationships between sub-periods. Thus, the current study fills the gap in the ongoing energy and economics literature.

Data and methods

Variables source and description

The empirical study evaluates the nexus between CO2 emissions and its drivers such as financial development, renewable energy, economic growth and environmental-related technologies in South Africa utilizing data between 1980 to 2020. The dependent is CO2, while the determinants of CO2 such as financial development, renewable energy, economic growth and related environmental technologies represent the independent variables. The precise and detailed information regarding CO2 and the regressors are presented in Table 1.

Table 1 Variables and measurement.
Full size table

The variables of investigation i.e., CO2, ETEC, GDP, FD, and REC are logged to ensure conformity to normal distribution in line with the studies of21,34. Following the studies of35,36, we formulate the following economic function:

$${\mathrm{LnCO}}_{2\mathrm{t}}={\beta }_{0}+{\beta }_{1}{GDP}_{t}+{\beta }_{2}{ETEC}_{t}+{\beta }_{3}{REC}_{t}+{\beta }_{4}{FD}_{t}+{\varepsilon }_{t}$$
(1)

where; \({\beta }_{0}\) denotes the constant, \({\beta }_{\mathrm{1,2},\mathrm{3,4}, and 5}\) represents the coefficients of the independent variables while \({\varepsilon }_{t}\) represents error term. CO2, GDP, ETEC. REC and FD stand for carbon emissions, economic growth, environmental-related technologies, renewable energy and financial development.

Table 2 presents the series descriptive statistics. The mean value of GDP (5301.2) is the highest, and it ranges from 4269.7 to 6263.1, REC (181.46) which ranges from 10.152 to 737.05, ETEC (10.404), which ranges between 4.3000 and 16.920, CO2 (8.4707) ranges from 7.2155 to 9.7892. The skewness value shows the variables are skewed positively, while the kurtosis value indicates that the variables are platykurtic with REC exemption which is leptokurtic. Figure 1 also presents additional information regarding the variables.

Table 2 Descriptive statistics.
Full size table
Figure 1
figure 1

Box Plot.

Full size image

Estimation strategies

The ARDL model established by37 was employed to evaluate the long-term connectivity between the variables. As a result, the model posits as follows;

$$\Delta Ln{CO}_{2t}={\sigma }_{0}+{\sigma }_{1}{\sum }_{i=1}^{p}\Delta L{nCO}_{2t-1}+{\sigma }_{2}{\sum }_{i=1}^{p}\Delta {LnGDP}_{t-1}+{\sigma }_{3}{\sum }_{i=1}^{p}\Delta Ln{ETEC}_{t-1}+{\sigma }_{4}{\sum }_{i=1}^{p}\Delta Ln{REC}_{t-1}+{\sigma }_{5}{\sum }_{i=1}^{p}\Delta Ln{FD}_{t-1}+{\beta }_{1}\Delta Ln{CO}_{2t-1}+{\beta }_{1}\Delta Ln{GDP}_{t-1}+{\beta }_{3}\Delta LnETE{C}_{t-1}+{\beta }_{4}\Delta Ln{REC}_{t-1}+{\beta }_{5}\Delta Ln{FD}_{t-1}+{\varepsilon }_{t}$$
(2)

Equation (2) represents the ARDL test's unrestricted ECM. The short-run coefficients shown by \({\sigma }_{1, 2, 3, 4 and 5}\). Moreover, \({\beta }_{1, 2, 3, 4 and 5}\) illustrates the coefficients of the long-run. Also, the long-run cointegration is estimated using the bound test F-statistic value for the Ho hypothesis of "no cointegration," which is denoted by \(\mathrm{H}0:{\upbeta }_{1}={\upbeta }_{2}={\upbeta }_{3}={\upbeta }_{4}={\upbeta }_{5}=0\). Furthermore, the null hypothesis is rejected when the bound test F-statistic value surpasses both the higher and lower bound values. Conversely, the Ho hypothesis is deemed valid if the value is smaller. Furthermore, \(t-1\) denotes the optimal lag for each variable as calculated by AIC.

The association between the variables is evaluated utilizing the error correction model. To ascertain the short-term characteristics, the lagged terms for each individual coefficient are also used. Furthermore, the error correction term (ECT) enables the collection of information on the long-term dynamics.

$$\Delta Ln{CO}_{2t}={\sigma }_{0}+{\sigma }_{1}{\sum }_{i=1}^{p}\Delta L{nCO}_{2t-1}+{\sigma }_{2}{\sum }_{i=1}^{p}\Delta {LnGDP}_{t-1}+{\sigma }_{3}{\sum }_{i=1}^{p}\Delta Ln{ETEC}_{t-1}+{\sigma }_{4}{\sum }_{i=1}^{p}\Delta Ln{REC}_{t-1}+{\sigma }_{5}{\sum }_{i=1}^{p}\Delta Ln{FD}_{t-1}+\varphi EC{T}_{t-1}+{\varepsilon }_{t}$$
(3)

The ECT shows short-term alterations while also taking the rate of return to equilibrium into account. Typically, the value of the ECT should range from minus 1 to 0. Furthermore, it ought to be significant and negative. Moreover, the models’ stability is tested utilizing CUSUM and CUSUMSQ.

Moreover, in line with the studies of38,39,40, we employed the time-varying causality test. Unlike the conventional causality tests such as17,18 causality tests, this technique, which uses a bootstrap** strategy to account for small-sample bias, enables us to identify any structural changes in the framework as well as the progression of causal interrelationships between sub-periods. The flow of the study analysis is presented in Fig. 2.

Figure 2
figure 2

Flow of the study.

Full size image

Findings and discussion

Preliminary tests

The study evaluates the stationarity features of the series as an initial test (see Table 3). The PP and ADF results documented similar results, showing non-stationarity as level with the exemption of ETEC; Nonetheless, it became stationary after the first difference was taken. In summary, the study series are stationary at mixed order i.e. I(1) and 1(1).

Table 3 Stationarity test results.
Full size table

Result of cointegration

Next the study used both ARDL bounds (see Table 4) test and Bayer–Hanck (see Table 5) to capture the long-run dynamics between CO2 and its drivers. The results from the bounds test disclosed evidence of long-run dynamics between CO2 and its drivers. Moreover, this result is also validated by the Bayer–Hanck Results cointegration. Thus, Ho hypothesis of “cointegration does not exist” is dismissed at 5% as reported by both Bayer–Hanck and ARDL Bounds tests cointegration.

Table 4 Bounds test.
Full size table
Table 5 Bayer–Hanck results.
Full size table

Results of ARDL

The study explored the drivers of carbon emissions for the case of South Africa using the ARDL, which can identify both short and long run interrelationships (see Table 6). CO2 emissions are positively impacted by economic growth in the long-and-short run. Specifically, a 1% upsurge in GDP contributes to the intensification of CO2 by 1.128%\(\sim\) longrun and 1.1827 \(\sim\) short-run. These reinforced the emissions-driven economic growth in South Africa. South Africa is an emerging economy; as a result; they pay close attention to economic progress while little/less attention is assigned to environmental sustainability. It is clear that the growth trajectory of South Africa's economy is not sustainable. Thus, achieving the SDGs will become easier if policy realignment is considered. The studies of14,20,41 also reported similar results.

Table 6 ARDL results.
Full size table

The study also uncovers the negative nexus between environmental-related technologies and CO2 emissions in both short and long-run. Specifically, 0.0772%\(\sim\) longrun and 0.0474 \(\sim\) short run decrease in CO2 emissions is caused by a 1% upsurge in South Africa's environmental-related technologies. This shows that related environmental technologies are essential for the decline in CO2 emissions, which, as a result, leads to ecological sustainability. As stated by42, several nations are investing in environmental related technologies because it is clean and sustainable. Moreover, environmental-related technologies are essential for the energy transition. Thus, South Africa is on the right path toward attaining SDG 7. This result affirmed the studies of19,43,44. However, the study of45 reported different results by establishing positive interrelationships.

Moreover, renewable energy impacts CO2 emissions negatively, though the impact is insignificant. This is a significant issue in the case of South Africa. According to the Paris accord, COP26 and COP27, renewable energy has been highlighted as the solution for achieving sustainable sustainability. Also, renewable energy is secure and long-lasting enough to benefit the environment without slowing down the pace of economic progress. In addition to being sufficient and environmentally benign, renewable energy sources also lessen CO2 emissions. This shows that South Africa needs to move towards energy transition swiftly. Moreover, achieving carbon neutrality target is in jeopardy in the case of South Africa. Consequently, policymakers in South Africa must readjust their policies towards achieving SDGs 6 and 7 goals. Our findings align with46 studies and47 for China, who highlighted an insignificant nexus between CO2 and renewable energy. However35,48,49 contradict our studies by establishing the declining emissions role of renewable energy.

The study also established the CO2 emissions decreasing role of financial development. In specific, the 0.1291%\(\sim\) longrun and 0.2813 \(\sim\) short run decrease in CO2 emissions is caused by a 1% upsurge in financial development in South Africa. The result shows that the financial system of South Africa is mature. The outcome suggests that South Africa’s financial sector embraces sustainable initiatives and prioritizes ecological sustainability. Considerable financial development that addresses ecological concerns helps to reduce CO2 emissions. This result complies with the studies of50,51. However, this study contradicts23,52, who highlighted that with low-interest loans, financial development increases the spending power of the general public. This enables people to buy opulent goods like houses, cars, and air conditioners, all of which put a burden on the environment.

Lastly, as expected, the ECT is negative and significant statistically which shows that an earlier adjustment could be rectified in the subsequent periods. Figure 3 shows the summary of the ARDL results.

Figure 3
figure 3

Summary of findings.

Full size image

Diagnostic test results

The study proceeds to the diagnostics tests by evaluating the reliability of the model. Table 7 presents the result of these tests. The results show no evidence of serial correlation issue; residuals are distributed normally, no issue of misspecification, and no problem of heteroskedasticity. Moreover, the CUSUM (see Fig. 4a) and CUSUM of square (see Fig. 4b) support the stability of the model.

Table 7 Diagnostics tests.
Full size table
Figure 4
figure 4

(a) CUSUM, (b) CUSUM of square.

Full size image

Time-varying causality test outcomes

Unlike prior studies that employed the conventional causality tests such as17,18 causality tests, we utilized the time-varying causality, which uses a bootstrap** strategy to account for small-sample bias, enables us to identify any structural changes in the framework as well as the progression of causal interrelationships between sub-periods. Figure 5 presents the time-varying causality results between CO2 emissions and the regressors. In Fig. 5, the thick black line depicts the 10% level of significance.

Figure 5
figure 5

(a) Causality between economic growth and CO2, (b) Causality between renewable energy and CO2, (c) Causality between related environmental technologies and CO2, (d) Causality between financial development and CO2.

Full size image

Figure 5a presents the causality between CO2 and GDP. The results show dismissed Ho hypothesis of "no causality" in 2007. On the flip side, the null hypothesis of "no causality" from CO2 emissions to GDP was dismissed in 1990, 2000, 2004–2006, 2008 and 2019. In summary, feedback causality exists between GDP and CO2 as shown by the time-varying causality. Figure 5b unveil causality between CO2 and renewable energy. Evidence of causality from renewable to CO2 is noticed from 2003–2004 while causality from CO2 to renewable energy is captured from 2001 to 2003. Figure 5c shows causality between environmental-related technologies and CO2. Specifically, causality surfaced from environmental technologies to CO2 from 1990–1992, 2003, 2011–2012, 2017 and 2019. Conversely, causality emerged from CO2 to environmental technologies from 2000–2001, 2004–2006 and 2015. Lastly, Fig. 5d display the causality between financial development and CO2. The Ho hypothesis of “no causality” is refuted in 2000 and 2006–2008 from financial development to CO2. On the flipside, Ho hypothesis of “no causality” from CO2 to financial development was refuted in 2014.

Conclusion and policy remarks

The empirical research pursues to evaluate the nexus between CO2 emissions and its drivers-financial development, renewable energy, economic growth and environmental-related technologies in South Africa utilizing data between 1980 to 2020. This investigation seeks to uncover this association more extensively, meticulously, and analytically to contribute to South Africa's CO2 reduction measures. In doing so, we employed autoregressive distributed lag (ARDL) and time-varying causality to evaluate these connections. The results from the ARDL show that financial development and environmental technologies lessen CO2 emissions while economic progress intensifies CO2 emissions. Surprisingly, renewable energy does not mitigate CO2 emissions. Furthermore, the time-varying causality shows that all the dependent variables can forecast CO2 emissions at sub-periods.

The findings of our study have a variety of significant policy recommendations. Environmental-related technologies help to reduce carbon emissions. Hence, the South African government should emphasize ecological innovations and acknowledge the significance of environmental-related technologies in lowering CO2 emissions. To address a worsening of the ecosystem brought on by CO2 emissions, governments and policymakers in South Africa must develop measures to increase investment in eco-friendly technologies. Therefore, the South African government should launch new research and development initiatives in environmentally friendly technologies, and authorities should collaborate with industry to create cutting-edge initiatives to combat the largest damaging factor—CO2 emissions from business operations. Furthermore, investors should be trained to invest in companies that are making significant efforts to reduce their adverse environmental impact by implementing eco-friendly technologies. Moreover, strategies should be developed and implemented to promote the corporate sector to take such actions.

South Africa relies far too heavily on fossil fuels. Resource and pollution taxes are two examples of carbon-emitting resource levies incentivizing consumers and companies to use renewable energy sources. Investors and customers must favor eco-friendly goods and services. Moreover, energy sources that emit CO2 should be substituted with renewables. Additionally, since diesel cars generate a lot of CO2 and fine particulates, high vehicle registration and circulation costs can prevent people from buying them. Furthermore, South Africa should shift to electromobility in the automobile industry and other transportation methods. These measures will not only help South Africa create a sustainable ecosystem but will also aid in transforming the nation's economy in favor of the environment.

The present study, like many others, has constraints. Due to data restrictions, we could not incorporate several other relevant drivers in our research, such as globalization, economic complexity, government stability, political risk, human capital, etc. Thus, future studies should incorporate those variables to assist the nation’s policymakers.