Unraveling the environmental consequences of trade openness in South Africa: a novel approach using ARDL modeling

Free trade and environmental sustainability are currently top economic and environmental research priorities. While numerous theories connect trade openness with environmental quality, empirical evidence often fails to support these claims. Using data spanning from 1980 to 2020, our study examines the effect of trade openness on CO2 emissions in South Africa. By employing a novel ARDL modeling framework, our analysis confirms the presence of the Environmental Kuznets Curve (EKC) hypothesis in South Africa. Our findings reveal that while GDP square enhances environmental quality, trade openness and economic growth have a degrading effect over time. Additionally, the study identifies that energy consumption, FDI, and value-added activities all contribute to environmental degradation. Findings also highlights the influence of institutional quality on the environment, demonstrating that political stability and control of corruption lead to increased CO2 emissions, while the rule of law reduces CO2 emissions. The research suggested that the potential of green economies should be leveraged in developing renewable energy, sustainable development, the recycling industry, and green financing sectors. A shift in economic activity in this direction will thus foster long-term economic growth and sustainable development.


Introduction
To achieve long-term growth, a sustainable environment, and development, it is essential to possess an in-depth view of the environment in which we interact.This is because, as human activity increased, the conflict between environmental quality and sustainability became a global issue.Due to poverty, a lack of industrialization, economic stagnation, and a lack of environmental concerns, Africa does not experience the same level of environmental degradation and hazards as Europe and North America.As a result of the continent's relentless pursuit of growth and development through technological and industrial advancement, environmental problems in Africa are becoming more serious.But if people want to improve the world's current structure or make it lovely for habitation, environmental sustainability must be combined with development ideals.Even though it is debatable whether humans have the capacity to meet specific environmental needs, deliberate efforts must be made to maximize the severe environmental effects of social actions.
Nonetheless, environmental issues is a significant concern and has draw much interest from economists and researchers.Khan et al (2022) stated that due to the ongoing raise in CO 2 emissions, countries are experiencing serious issues with global warming.According to Shahbaz et al (2013), environmental degradation is no longer limited to industrialized countries, and the accumulation of greenhouse gas (GHG) emissions significantly impacts every nation, regardless of where the emissions.Zerbo (2017) argued that human-caused environmental effects have led to disasters, harming natural resources and human lives.However, the major GHG emitters, such as the BRICS, OECD, and the U.S, according to Shahbaz et al (2013) are responsible for reducing global carbon emissions to combat the effects of global warming.The participation of these major polluters is critical to the success of global efforts to address environmental issues.Research by Ding et al (2021) deter sustainable development in the country.The research is expected to provide a solution to the problem by analyzing the intricate linkage between trade openness and environmental degradation, specifically focusing on CO 2 emissions.As mentioned above, the unique methodologies and conceptual background of the study are anticipated to generate new and interesting findings that contribute to improving our understanding of the relationship between trade liberalization and the environment.The choice of sample in the study also complements the problem by focusing on South Africa, a country that has seen significant growth in trade openness but has received less attention regarding the environmental consequences of its trade policies.While the results and policies derived from the study may be specific to the context of South Africa, the findings can potentially be generalized to other countries facing similar challenges related to trade openness and environmental degradation.The methodologies and analytical approach employed in the study may offer valuable insights that can be adapted and applied in other regions to inform policy decisions aimed at balancing economic growth with environmental sustainability.
However, the remaining work is divided as follows: section 2 contains literature review while section 3 contains theoretical framework.Section 4 presented methods and section 5 contains results, while section 6 and 7 provided a discussion and conclusion.

Literature review
Based on the theoretical foundations developed by Copeland and Taylor (1994) and Grossman and Krueger (1991), the study examines the effects of trade openness on the environment.However, following the work of Grossman and Krueger (1991), a multitude of studies have enriched the discourse on the link between economic growth and CO 2 emissions.Shahbaz and Sinha (2019) have adeptly synthesized these studies through comprehensive literature reviews.Nevertheless, conflicting conclusions persist among these studies.While certain studies, such as those by Apergis and Ozturk (2015) Murshed and Dao (2020), Ibrahim et al (2023) and Ghazouani, Maktouf 2024) have utilized panel data, others like Shahbaz et al (2014), Seker et al (2015), Usman et al (2019) and Awan et al (2022) have employed time series data to support the presence of an inverted U-shaped relationship.Conversely, research by Adu and Denkyirah (2019), Farhani and Ozturk (2015), Ben Jebli and Ben Youssef (2015), and Amri (2018) challenge this notion.Moreover, investigations into the nexus between economic growth and CO 2 emissions transcend the Environmental Kuznets Curve (EKC) theory.Some studies opt for a linear approach to explore whether economic growth leads to a proportional rise or fall in CO 2 emissions.These studies, typically grounded in time series data, offer a longitudinal perspective, while others leverage panel data to capture shared patterns and inter-regional disparities.For instance, Magazzino (2016) illustrates that economic growth contributes to CO 2 emissions in nations like Armenia, Azerbaijan, Georgia, and Turkey.
However, numerous studies on the connection between trade and CO 2 emissions have shown contradictory and unclear results.Certain studies indicate that trade liberalization enhances environmental quality via multiple channels (Khan et al 2020, Ding et al 2021, Ibrahim and Ajide 2021).Additional research has demonstrated that trade liberalization has no effect on the environment (Oh andIqbal 2018, Gulistan et al 2020) while Ghazouani and Maktouf (2024) found a unidirectional link between trade openness and CO 2 emissions in oil-exporting countries.Research by Ding et al (2021) demonstrates that increased trade openness in the G-7 economies is linked to improved environmental quality using CS-ARDL and augmented AMG techniques.Ibrahim and Ajide (2021) found that trade openness lowers environmental degradation in G-20 nations using CCEMG and MG. .According to an empirical study by Khan et al (2021), Pakistan's environment worsens when trade is open.This empirical conclusion is corroborated by Ibrahim and Ajide's (2021) findings.Azam et al (2020) claim that trade openness, financial development, and economic expansion are just a few of the important channels via which institutions affect carbon emissions.Van Tran (2020) found that trade openness negatively affects environmental conditions in 66 emerging and developing economies.Khan and Ozturk (2021) claim that trade openness raised CO 2 emissions in 88 developing nations.Ali et al (2020) claim that the opening of the OIC countries to the world product market has a detrimental effect on their environmental conditions.Aydin and Turan's (2020) findings for China and India add to this findings for the years 1996-2016.Using OLS regression, fixed effects, and the GMM, Khan et al (2022) discovered that while trade openness, the use of renewable energy, and foreign direct investment have a negative relationship with carbon emissions, most institutional quality indicators have a significant impact on environmental sustainability.
However, studies regarding South Africa's situation are conflicting and diverse (Shahbaz et al 2013, Zerbo 2015, 2017, Inglesi-Lotz 2018, Mapapu and Phiri 2018, Udeagha and Ngepah 2019).Even though trade has been shown to improve environment in the short term, Udeagha and Ngepah (2019) argue that access to international goods markets harms South Africa's environment over time, resulting in a significant asymmetry between CO 2 emissions and trade openness.Udeagha and Ngepah (2022) emphasized the significant negative effects of coal-dependent energy systems on the quality of environment in South Africa in their study.In Aydin and Turan (2020) study, trade improves environmental quality by minimizing the increasing rate of energy pollution.Adams et al (2019) emphasized the need for green fiscal measures to mitigate the adverse environmental effects that could result from more open trade.By making pollution an expensive proposition for industries, these green fiscal changes can act as financial obstacles, promoting the adoption of cleaner technologies.
Frontera et al (2021) stated that the correlation between trade openness and environmental has continued to generate substantial debates, as these two criteria are crucial for sustainable development.Fatima et al (2019) highlighted that while trade openness may increase energy needs, it may also pave the path for sustainability by fostering technology transfers and incorporating cleaner fuels into energy blends.According to Ali et al (2017), technical innovations frequently traverse international commerce channels, regardless of how they may affect the environment.Unfortunately, some of this imported technology is usually produced in resource-intensive and high-emission factories in underdeveloped countries.Zhang et al (2021) evaluated the effects of trade liberalization and prospective goals for economic development in their study, and their results revealed that these factors significantly influence the energy sector's development in a direction that is more environmentally sustainable.Ebenyi and Kalu (2020) suggested that a multidimensional strategy is needed to utilize trade's maximal potential for sustainable development.This strategy must include not only technological innovation and economic laws but also effective governance and public awareness.

Theoretical framework
Numerous studies examining the relationship between economic growth, trade liberalization and environmental sustainability in various countries and regions exhibit differences, as outlined in the literature review section.The number of research conducted in this field has surged in the last three decades following the adoption of the Montreal Protocol in 1987 and the UNFCCC in 1992.The emphasis on environmental preservation was further reinforced by the more recent and comprehensive Paris Agreement of 2015.Here, we discuss two commonly utilized theories that aim to elucidate the link between trade openness and environmental degradation: the Environmental Kuznets Curve (EKC) hypothesis and the Pollution Haven hypothesis (PHH).
First the Environmental Kuznets Curve (EKC) hypothesis was first introduced to the environmental field by Grossman and Krueger (1991), building upon Kuznets (1955) theoretical framework.The hypothesis posits a curvilinear relationship between economic growth and environmental degradation, taking the shape of an inverted U curve.This suggests that as economic activities expand, there is a heightened impact on the environment due to increased pollutant emissions and the extensive or intensive utilization of natural resources linked to the adoption of specific production technologies.In essence, the EKC hypothesis suggests that environmental deterioration escalates during initial growth stages, reaches a tipping point, and then declines as income continues to rise.One contributing factor to the rise in pollution during early economic growth phases is the hesitance of policymakers to enforce stringent eco-friendly regulations on production out of concern for deterring potential investors, particularly foreign direct investments, which are crucial in the early phases of economic advancement.The priority placed on domestic investment and economic expansion outweighs the emphasis on environmentally sustainable production processes within the economy.
The theory proposes a transition in policy towards environmentally sustainable production methods once the economy reaches a certain level of economic development.This typically occurs when the income levels of the country reach a specific threshold.At this stage of economic progress, the country's focus shifts from simply increasing production to promoting improved quality consumption, which includes factors like good health and a clean environment.Policymakers also introduce and rigorously enforce stringent regulations necessary to uphold such environmental standards, requiring all production processes to adhere to these regulations.Considering South Africa as the case study, we anticipate a positive relations between economic growth and CO 2 emissions since these nation is in the developmental and emerging stages, necessitating increased foreign capital and investment inflows.We also anticipate a shift in this relationship after reaching a certain level of development, thereby confirming the Environmental Kuznets Curve (EKC) hypothesis.Numerous studies have been conducted in diverse contexts to verify the existence or absence of the EKC hypothesis, yielding inconclusive results.For instance, Gyamfi et al (2021), Tenaw and Beyene (2021), and Weimin et al (2022) identified evidence supporting the EKC hypothesis in their research, while Allard et al (2018), Awan and Azam (2021) did not find evidence of its presence but noted an N-shaped relationship between environmental pollution and economic growth.Figure 1 illustrates the functioning of the EKC hypothesis by depicting the varying levels of economic development in an economy and the corresponding pollution levels produced over time.
In figure 1, an economy operating at the pre-industrial phase exhibits high levels of environmental pollution and demonstrates minimal concern for environmental sustainability.During this stage, the primary objectives revolve around boosting investments and fostering economic growth.Upon transitioning to the industrial phase, the economy experiences a decrease in emission rates compared to the pre-industrial era.Environmental pollution reaches its peak at this juncture, commonly referred to as the turning point, before gradually declining.Pollution levels rise leading up to the turning point but diminish thereafter.Consequently, in the post-industrial phase, economies prioritize citizen welfare, quality of life, and consumption, leading to the strict enforcement of environmental regulations to safeguard their populace.
Secondly, the Pollution Haven Hypothesis (PHH), also known as the pollution haven effect, posits that industries that generate pollution will move their operations to areas with less stringent environmental regulations.This theory suggests that regions with lenient environmental laws (referred to as pollution havens) will attract industries relocating from areas, particularly industrialized countries, with stricter regulations.This relocation is driven by the desire of industries to lower production costs, as stringent environmental regulations increase the overall cost of production.While the Heckscher-Ohlin model argued that regions with relaxed regulations would hold a comparative advantage in producing specific goods, this scenario is rarely observed in reality due to the significant implicit costs associated with environmental damage caused by production activities.Polluting industries tend to segment their production processes and shift the pollution-intensive stages to developing countries with less stringent regulations.Consequently, limited benefits are realized from such production practices, primarily in the form of informal labor opportunities.These financially robust industries often wield influence through lobbying efforts and pay minimal taxes, resulting in societal losses that primarily benefit few individuals, typically the political elite.The concept was first formulated by Walter and Ugelow (1979) and subsequently supported by Baumol et al (1988).A notable example is the extractive industry in many resource-rich developing nations, where environmental pollution occurs locally, while the raw materials are exported and processed in developed countries.However, our study will integrates both the EKC hypothesis and Pollution Haven Hypothesis (PHH) to analyze how trade openness affect environmental degradation in South Africa.By combining these two theories, the research aims to provide a comprehensive understanding of how trade openness respond to varying levels of environmental regulations in South Africa.This dual theoretical framework also allows for a nuanced examination of other factors influencing environmental degradation South Africa.
3.1.Methods 3.1.1.Functional form However, following the robust empirical approach widely used in earlier research, the impact of trade openness on CO 2 emission in South Africa is revisited using the EKC hypothesis framework.Thus, in its standard form, following Udeagha and Breitenbach (2021) According to Ling et al (2015) and Cole and Elliott (2003), the technical effect of environmental law implementation and enforcement, as well as people's preference for a healthy and sustainable environment, improved environmental quality while the scale effect deteriorated as income increased.However, it is assumed that the theoretical expectation that Φ > 0 and β < 0 is valid for EKC theory to arise based on this assertion.
As a result, the control variables are foreign direct investment (FDI), use of energy (ENG), technical innovation (TECH), value-added (PGDP), population growth (POP), and Institutional quality (INT) (which include 3 blocs of political stability, rule of law, and corruption index).As a result of combining these variables with trade openness (TRO), our equation (2) can be stated as follows: β 1 -β9 represent the coefficients of independent variables that capture different elasticities of the variables under study while μ t is the stochastic error term.The variables a in natural logs.

Data and variables used
This analysis made use of annual data from 1990 to 2020.The CO 2 is the dependent variable, whereas the explanatory variable is trade openness (TRO) which was sourced from World Bank (2021).In South Africa, GDP square, which captures the technical effect, was used to ascertain the existence of EKC theory, whereas the scale effect is a proxy for GDP.Controllable variables, on the other hand, include value-added, energy usage, technological innovation, FDI, population growth (POP), and Institutional quality (INT) (which include 3 blocs of political stability, rule of law, corruption index).
The composite trade intensity (CTI) technique built by Squalli and Wilson (2011) is utilized in this research to assess trade openness.This was done to get reliable statistics on GDP trade share and trade volume in relation to world trade.This approach to evaluating trade openness will avoid and fix the constraints of previous research's traditional trade intensity (TI).
Therefore, the calculation of CTI is as follows: Where i stands for South Africa, and j denotes for the country's trading partners.The first component of equation ( 4) explain share of global trade, while the second component denotes trade share of South Africa.Table 1 reports summary of variables explanations.

Narayan and popp structural break unit root test
The long-run correlation between the variables is analyzed using the bound test in this paper.Here we present the ARDL bound testing techniques: Where i denote the first difference for each of these variables: lnCO 2 , lnSE, lnTE, lnTRO, lnENG, lnFDI, lnTECH, lnPGDP, lnPOP and lnINT.g and Φ denote the short-run and long-run coefficients, respectively, and t-i is the selected optimal lags based on SBIC.Variables that are co-integrated provide short and long-run estimations to the ARDL model.The long run relationship is tested for the null hypothesis of (H 0 : To ascertain the evidence of co-integration among the variables, the F-statistic is employed.Evidence of cointegration is present if F-statistic value is higher than the upper bound.The test is inconclusive if it falls between the bounds, and there is no co-integration if it is less than the lower bound.The long-run ARDL model can be calculated using the following method if co-integration is demonstrated: The long-run coefficients in equation ( 6) is represented by p.This study uses the SBIC in order to determine the best lags.Nonetheless, the short run ARDL ECT model is: Equation (7) explain the variables' short-run unpredictability, while the error correction term (ECT) reflects the speed of readjusting any potential disequilibrium, with estimated values ranging between −1 and 0. Diagnostic checks are also employed to ensure the model's stability.The models are validated using the Ramsey RESET test; serial correlation is assessed using the Breusch-Godfrey LM test; heteroscedasticity is assessed using the Breusch-Pagan-Godfrey test and the ARCH test; the normality of the estimated residuals is assessed using the Jarque-Bera test; and structural stability is investigated using CUSUM and CUSUMSQ.

Frequency domain causality test
This research investigates the causality between the variables of interest by employing the frequency domain causality (FDC) method.Unlike the conventional Granger causality method, the method captures causation across short, medium, and long terms among the studied variables, and enables the forecasting of the response variable at a specific time frequency.

Summary statistics
Our study extensively examined the summary statistics of the employed variables, presented in table 2. Kurtosis indicates the peak, and the Jarque-Bera test assesses the normal distribution of the data series.

Unit root test
Table 3 demonstrates that according to the DF-GLS, PP, ADF, and KPSS tests, the data exhibit stationarity after the first difference I(1).This suggests that all the series are I(1).However, the testing methods utilized in this study enable the investigation of two structural breaks in the variables, with the findings presented in the right section of table 3. The results indicate that the unit roots (Ho) are not valid, implying that all the data are integrated at order one I(1), allowing for the application of the dynamic ARDL bound testing approach to all data series.

Lag selection results
Table 4 presents the outcomes of different tests employed for lag selection.Previous research consistently indicates that HQ, AIC, and SIC are commonly utilized for this purpose.In this study, SIC is specifically utilized for lag selection, and the findings suggest that lag 1 is the most suitable for our model.

Cointegration test results
The cointegration test result are displayed in table 5 based on Kripfganz and Schneider (2018) method.The F and t-statistics exceed the critical values at all significance levels, leading to the rejection of the null hypothesis.This suggests that the variables are cointegrated.

Diagnostic statistics tests
Table 6 displays the diagnostic tests result.According to the result, the Breusch-Godfrey LM test shows that there is no serial correlation in the model.Ramsey's RESET test reveals that the model is appropriately specified with no misspecifications.The assessment of heteroscedasticity using the Breusch-Pagan-Godfrey and ARCH tests indicates that it is relatively insignificant for the study.Lastly, the Jarque-Bera test results demonstrate that the model's residuals exhibit a normal distribution.

Results of dynamic ARDL simulations model
Table 7 presents the outcomes of the dynamic ARDL simulation models, indicating that the scale effect has a significant positive correlation with CO 2 emissions, while the technical effect has a significant negative correlation.Specifically, the results suggest that a 1% increase in GDP measured by the scale effect leads to a The results also suggest a positive association between trade openness and CO 2 emissions, with a 1% rise in trade openness leads to 0.30% rise in CO 2 emissions.This supports Lopez's (1994) claim that energy-intensive firms, which prioritize trade liberalization, are sources of pollution, as well as Taylor's (2005) pollution haven hypothesis.These findings are consistent with the environmental predicament in Pakistan, as claimed by Khan et al (2021a).However, they contradict the observations of Van Tran (2020) that trade openness worsens the state of the environment in 66 emerging economies.
Additionally, the results suggest that there is a significant positive association between use of energy and CO 2 emissions.This implies that increased CO 2 emissions are primarily due to use of energy, with a 1% rise in energy use resulting in a 0.20% rise in CO 2 emissions.South Africa is the world's seventh-largest coal-consuming nation, as reported by the World Bank, and while this is necessary for industrial growth and economic development, it also contributes significantly to environmental damages.This finding is consistent with Adebayo et al (2021) discovery that energy consumption in South Korea contributes to CO 2 emissions.However, some studies contradict these findings (Baye et al 2021, He et al 2021).
Our results indicate that foreign direct investment in South Africa has a positive relation with CO 2 emissions, leading to environmental degradation due to the country's comparative advantage in exporting and producing goods.Specifically, our result shows that a 1% rise in FDI leads to 0.63% rise in CO 2 emissions, supporting Copeland and Taylor's (2013) argument that polluting industries relocate to underdeveloped countries, worsening environmental degradation.This is consistent with Abdouli and Hammami's (2017) observation that FDI has significantly raise CO 2 emissions in the MENA region, as well as the pollution haven hypothesis.This finding is consistent with Haisheng et al (2005) The study revealed that technological innovation reduces CO 2 emissions, with a 1% rise in technical innovation leading to a 0.39% decrease in CO 2 emissions.This suggests that technological innovation in South Africa contributes to lowering greenhouse gas emissions.Similar conclusions were drawn by Sohag et al (2015), indicating innovation facilitates more efficient energy use and reduced greenhouse gas emissions.These findings  Furthermore, the study indicates that value-added has a long-term positive correlation with CO 2 emissions, suggesting that the growth of South Africa's manufacturing sector is closely linked to the rise in CO 2 emissions.Udeagha and Ngepah (2021) findings are consistent with our result, who also found that expanding industrial sectors contribute to increased CO 2 emissions.However, other studies, including those by Ewing and Rong (2008), Zhang and Lin (2012), suggest that the expansion of the industrial sector actually reduces CO 2 emissions.
Moreover, the study reveal that there is a positive correlation between population and CO 2 emissions.This suggests that the rising population in South Africa, along with increased per capita energy consumption, has accelerated environmental degradation.These findings are inline with the work of Markham (2008), Ibrahim et al (2023), and Khan et al (2022) that environmental degradation is primarily caused by changes in consumption habits rather than population growth.
The model includes three institutional quality variables: political stability, rule of law, and corruption control, in order to comprehensively examine the role of institutions in trade policy while considering environmental quality.These variables represent the political and judicial systems of South Africa.The result reveal that a 1% improvement in the rule of law will reduce CO 2 emissions by 0.04%.This indicates that institutions play an important role in improving the environment, particularly when the rule of law is strengthened and enforced.However, political stability as well as the control of corruption have a positive link with emissions, suggesting that these factors are weak and have a negative impact on environment.These results align with Khan et al (2022) result.
The ECT is statistically significant and negative values of −0.8584 indicate a long-run relation between the variables.This value suggests that 86% of short-term disequilibrium will be corrected and return to equilibrium in the long term.
Moreover, the study predicts that the variables under study may cause a 10% fluctuation in South Africa's CO 2 emissions.Figure 2 illustrates the link between economic growth and emissions as an impulsive response, depicting the scale effect transitions and their effects on emissions.If the scale effect increases by 10%, economic growth had a positive long-term impact on CO 2 emissions, whereas a decrease of 10% will have a negative longterm impact.However, the former has a greater impact than the latter.This suggests that rapid economic expansion has negatively affected environmental quality, while slower economic growth has a positively affected it in both the long and short run.
Figure 3 shows the impact of technical effect on CO 2 .It reveals that a 10% raise in the technical effect has a negative short-and long-term impact on CO 2 , while a 10% reduction has a positive effect.In essence, in South Africa, an increase in the technical effect benefits the environment, while a decrease has negative effects in both the short and long run.
Figure 4 illustrates the link between trade openness and CO 2 .It demonstrates that a 10% raise in trade initially has a negative effect on CO 2 in the short term but a positive effect in the long term.
Conversely, a 10% decrease in trade openness has a short-term positive effect on CO 2 emissions but a negative long-term effect.This suggests that while increased openness benefits South Africa's environment in the short run, it has adverse long-term effects.On the other hand, reducing trade openness benefits the environment in the long run but negatively affects it in the short term.
Figure 5 indicates the connection between energy use and CO 2 .It shows that a 10% raise in energy use, raise CO 2 in both the short and long run, while a reduction in energy use has the opposite effect.This implies that rising energy usage has worsened South Africa's environmental degradation.
Figure 6 displays the impulse response graph for South Africa's CO 2 emissions and foreign direct investment (FDI).It demonstrates that a 10% rise in FDI has a significantly positive short-and long-run effect on CO 2 emissions, while a 10% reduction has negative short-and long-run effects.These results indicate that more FDI has the opposite effect of improving South Africa's environmental quality.
Figure 7 depicts the relationship between CO 2 emissions and technological innovation in South Africa through the impulse response graph.The graph demonstrates that a 10% increase in technical innovation has both short-and long-term negative impacts on CO 2 emissions.This suggests that greater technological innovation can enhance South Africa's environmental state, whereas decreased innovation has detrimental effects on the country's environment in both the short and long term.
The impulse response graph in figure 8 demonstrates the relationship between CO 2 emissions and industrial value-added in South Africa.It indicates that a 10% raise in industrial value-added has both short-and longterm positive effects on CO 2 emissions, while a 10% decline has negative effects.This implies that an increase in  industrial value-added worsens South Africa's environment in both the short and long term, while a decrease improves environmental quality.
For this purpose, CUSUM and CUSUMSQ measures are employed which are depicted in figures 9 and 10.According to rule thumb, if the plots remain within a critical bound level of 5%, it indicates that model parameters are consistent over time.The trends in figures 9 and 10 illustrate the model's behavior over time, and since both CUSUM and CUSUMSQ fall within their 5% confidence intervals, we can conclude that the model parameters are steady.

Robustness
The measure developed by Squalli and Wilson (2011), known as composite trade intensity (CTI), is compared to the more conventional trade intensity (TI) measure using two different statistical tests.First, the study compares the partial correlation coefficients between CTI and TI using different indicators of free trade and capital flows, and secondly, the study re-estimates the models with the TI measure to test and verify how much our findings depend on the particular measurements we employ, thus providing conclusive evidence for the hypothesis.In this context, the TI is calculated by calculating the ratio of the sum of exports and imports to the gross domestic product.

Correlations
First, we conduct a correlation analysis (see table 9) between the two measures and various indicators of capital flows and free trade policies to assess the reliability of CTI and TI.In this analysis, we utilize the number of newly registered businesses as a percentage of total registrations, taxes on international trade as a percentage of revenue, customs and other import duties as a percentage of tax revenue, and FDI inflows and outflows as a percentage of GDP, as suggested by Squalli and Wilson (2011), as alternatives for liberal trade policies.All of proxy measures are obtained from data in the World Bank's global development indicators.It is expected that customs and other import tariffs will show a negative correlation with CTI and TI measurements, while other indicators of liberal trade policies are expected to have positive associations with these measures.Customs and other import duties were shown to have a negative correlation with both the CTI and TI trade openness indicators (−0.63 versus −0.14, respectively; see table 9).However, there is a higher correlation with CTI than with TI.CTI likewise appears to have a greater link with international trade taxes than TI, with estimated values of −0.27 versus −0.18.CTI correlates positively with all of the other proxies used, whereas TI correlates positively with just FDI net inflows and the number of newly registered businesses.CTI has a stronger correlation with new business registrations than TI.However, the correlation between TI and FDI net inflows  appears to be larger than CTI, with estimates of 0.48 and 0.20 for CTI, respectively, while it has a negative correlation with FDI outflows.These results raise questions about the dependability and credibility of the TI measure of trade openness.

What degree of variation do the measurements account for in the results?
To evaluate the sensitivity of the results to the measurements employed to validate the claim that CTI is more effective than TI, we conduct a second robustness test.For this purpose, we re-evaluate our models using TI instead of CTI.Table 7 demonstrates that when TI is applied, the R 2 value reduces dramatically from 0.8893 to 0.5186, showing a considerable fall in explanatory variable strengths.When the TI measurement is employed, there is a notable rise in the root-mean-squared error (RMSE) of the models, increasing from 0.076 to 0.2368, as indicated in table 7.Moreover, the coefficient of trade openness, which is the study's variable of interest, substantially increases when TI is used instead of CTI.Additionally, it appears that the TI estimates may have overestimated the effect of openness on environmental quality in South Africa due to the considerable difference between their estimates and the CTI estimates.Furthermore, the outcomes are highly dependent on whether TI or CTI is employed.The long-run coefficient estimated for Technological Innovation (lnTECH) is not only statistically insignificant but also contrary to the expected outcome.As a result, it is evident from the aforementioned tests and evidence that the CTI measure offers a statistically superior estimation of the models.This raises doubts about the application of the TI measure and provides statistical evidence in favor of the use of the CTI measure in this investigation.

Discussions
The findings concerning economic growth and CO 2 emissions indicate that as economic growth increases, CO 2 emissions also increase.However, the significance of the square of GDP per capita suggests a decrease in carbon emissions as the square of GDP per capita increases.This discovery is consistent with the Environmental Kuznets Curve (EKC) theory, which proposes that economic development initially raises emissions until they reach a certain level, after which environmental quality begins to improve.This improvement can be accomplished through the implementation of environmental policies and regulations.Some studies propose that trade openness leads to higher carbon emissions and environmental degradation, despite claims by some scholars that trade enhances environmental quality through composition, trade, and technical effects.
Our findings also highlight the significant volume of trade activities dependent on fossil fuels for both transportation and production, illustrating how increased trade openness in South Africa contributes to a decline in environmental quality.This outcome supports the pollution haven hypothesis, which suggests that developing countries like South Africa have a comparative advantage in producing goods with high levels of pollutants, while wealthier nations excel in manufacturing clean goods.
Our findings indicate that despite energy being extensively utilized to drive economic growth in the country, it has an adverse effects on South Africa's environment.The nation's pollution levels are escalating due to the heightened demand for energy in production, consequently leading to a decline in environmental performance.South Africa is ranked as the seventh-largest coal-consuming country globally, as per a World Bank assessment, in order to support industry and foster economic progress.However, this also significantly contributes to environmental degradation.The outcomes we have obtained align with theories asserting that the utilization of renewable energy lower carbon emissions and consequently enhances the environment, while the use of nonrenewable energy leads to pollution and harms the environment.As clean energy sources replace filthy fossil fuels, using renewable energy gradually improves the environment relative to using nonrenewable energy.In contrast, renewable energy has a direct correlation with sustainable development due to its accessibility, economic advantages, positive health effects, and reduced difficulties associated with social and environmental problems.
The results of research indicate that FDI inflow and carbon emissions in South Africa are positively correlated.According to the findings from our study, technological emissions, and contaminated goods are transferred through foreign direct investment, thereby degrading the environment's quality of the country.The findings demonstrate that the entry of polluting firms into South Africa exacerbated the country's environmental degradation.Technical innovation reduces carbon emissions by enabling cost-effective methods of lowering carbon dioxide emissions and by maximizing energy efficiency based on the findings.This may contribute to the economy's shift toward the production and use of sustainable energy sources.Furthermore, our research revealed a strong correlation between the growth of South Africa's manufacturing sector and rising CO 2 emissions, indicating that the country's developing industrial sectors were the primary drivers of the country's rising CO 2 emissions.South Africa's growing population has increased per capita energy consumption and has accelerated environmental deterioration.In other words, the growing population exacerbates climate change by increasing environmental pressure on fossil fuels and other non-renewable energy sources, as a result of the country's dense population and continual rapid expansion in urban, suburban, and environmentally fragile regions.
Our research findings demonstrate that strict adherence to legal regulations enhances environmental quality in South Africa.Conversely, political stability and corruption control are associated with declining environmental standards in the country.Therefore, it is imperative to bolster institutional frameworks to enhance environmental conditions.This involves the development of environmental policies and the reinforcement of national laws and regulations.Stronger institutions reflect the presence of the rule of law and human existence, which are essential for maintaining economic freedom and market economy, ultimately leading to improved environmental quality.Robust institutions facilitate the enforcement of energy-related regulations and encourage the adoption of renewable energy technology.Collaboration among various institutions is crucial when implementing environmental legislation to safeguard the environment.
Our research, however, is limited to South Africa.Future research can look into this link by looking at Africa as a whole.Future research could delve into the pollution haven and pollution halo hypotheses within this context, expanding on our examination of the environmental Kuznets curve.While our study focused on regional data, future inquiries could compare the impact of these factors on carbon emissions in both developed and developing nations.Technological innovation has shown significant potential in lowering CO 2 emissions.Therefore, further studies on the adoption, dissemination, and impact of environmentally friendly technologies, as well as on technological exchange facilitated by international trade and investments, can guide policy decisions on technological direction and innovation systems.The evidence of trade liberalization and environmental sustainability indicates a complex correlation.A long-term exploration of this relationship can provide interesting findings.For instance, research could seek to understand what happens as societies implement trade liberalization policies over time and how this intersects with the state of environmental sustainability.Unpacking the relations between the economic growth brought about by trade and its impact on environmental sustainability can result in designing policies and strategies that maximize the benefits of trade while mitigating its environmental cost.

Economic implications of the finding
The economic intuitions of the results outlined above are significant and multifaceted.The positive relationship between economic growth and CO 2 emissions suggests that as economies expand, they tend to produce more greenhouse gases.This has implications for policymakers and businesses as they need to consider strategies to decouple economic growth from environmental degradation.The Environmental Kuznets Curve theory provides a framework for understanding that at a certain level of economic development, environmental quality can improve.This implies that there is a need for targeted policies and regulations to ensure that economic growth is sustainable and does not come at the cost of environmental degradation.
The findings indicating that trade openness can lead to higher carbon emissions and environmental degradation highlight the need for South Africa to consider the environmental impact of their trade activities.As the pollution haven hypothesis suggests that developing countries may be more likely to produce goods with high levels of pollutants, which can have negative implications for the environment, this underscores the importance of incorporating environmental considerations into the South Africa's trade policies and agreements to mitigate the adverse effects of trade on the environment.Also, the reliance on fossil fuels for energy production and transportation in South Africa is found to be the contributing factor to environmental degradation.This implies that as the country continues to use extensive energy to drive economic growth, there is a need to transition towards cleaner and renewable energy sources.This shift towards renewable energy will not only improves environmental quality but also has economic benefits in terms of creating new industries and jobs in the renewable energy sector.
The positive relationship between FDI inflow and carbon emissions in the results highlights the role of foreign investment in contributing to environmental degradation.Based on this result, it is essential for policymakers to regulate and monitor the environmental impact of foreign investment to ensure that it does not exacerbate environmental issues.This can be done by encouraging technological innovation and energy efficiency through FDI which can help reduce carbon emissions and promote sustainable development.Also, the importance of strong institutional frameworks in enhancing environmental quality cannot be overstated.Strict adherence to legal regulations, political stability, and corruption control are crucial for maintaining environmental standards.Strengthening institutions and enforcing environmental policies and regulations are essential for improving environmental conditions and promoting economic growth in a sustainable manner.However, addressing environmental challenges requires a holistic approach that considers the economic, social, and environmental dimensions of development.Policymakers, businesses, and stakeholders need to work together to implement policies and practices that promote sustainable development and protect the environment for future generations.

Political implications of the findings
The political implications of the research findings are significant and underscore the need for strategic policy interventions to address environmental challenges in South Africa.The positive relationship between economic growth and CO 2 emissions highlights the importance of implementing sustainable development strategies that prioritize environmental protection alongside economic progress.Our findings on the the Environmental Kuznets Curve suggests that as South Africa's GDP per capita increases, there is potential for a decrease in carbon emissions, indicating the possibility of achieving a balance between economic growth and environmental sustainability through targeted policies and regulations.
Our findings on impact of trade openness on CO 2 emissions emphasizes the need for environmental considerations to be integrated into trade policies.This is the increased trade activities dependent on fossil fuels contributed to environmental degradation which confirmed the pollution haven hypothesis.This finding, however, highlights the importance of regulating trade activities to mitigate their adverse environmental impacts and promote sustainable trade practices.Also, the reliance on fossil fuels for energy production and transportation in South Africa is shown to have negative effects on the environment, with the country being a significant coal consumer globally.This underscores the urgency of transitioning towards cleaner and renewable energy sources to improve environmental performance and reduce carbon emissions.Therefore, policy measures aimed at promoting renewable energy technologies and enhancing energy efficiency in this regard can play a crucial role in mitigating environmental degradation and fostering sustainable development.
The positive relationship between FDI inflow and carbon emissions from our results indicates the need for regulatory oversight to ensure that foreign investment does not exacerbate environmental degradation.Thus, strengthening institutional frameworks, developing environmental policies, and enforcing national laws and regulations are essential for enhancing environmental quality and promoting sustainable development.Furthermore, political stability, corruption control, and collaboration among institutions are crucial for effective implementation of environmental legislation and safeguarding the environment.However, the importance of proactive policy interventions and collaborative efforts among stakeholders and concerned to address environmental challenges in South Africa is necessary.By prioritizing environmental protection, implementing sustainable development strategies, and promoting renewable energy technologies, policymakers can work towards achieving a balance between economic growth and environmental sustainability for the benefit of current and future generations.

Conclusion
The paper examines the impacts of trade openness on South Africa's environmental degradation proxy by CO 2 emissions from 1980 to 2020 using Jordan and Philips (2018) newly built ARDL model.We are able to overcome the constraints of prior research by using this method to determine the positive and negative connections between the explanatory variables and CO 2 emissions in South Africa.In order to confirm the reliability of our findings, we took a robustness check using the frequency domain causality technique, a robust testing tool developed by Breitung and Candelon (2006).This method allows us to examine short-, medium-, and long-term causation among the variables of interest.We confirmed the asymptotic behavior and integration order of all variables included models through the unit root tests.Additionally, we utilized the structural break unit root test developed by Narayan & Popp to account for the potential impact of structural breaks on various macroeconomic variables, including CO 2 emissions and trade openness.Consequently, all tests suggested that the data series are integrated at order one (I(1)), without presence of higher-order integration (I(2)).
Based on our research, we found that in South Africa, the scale effect exacerbates environmental degradation, while the technique effect contributes to its improvement, aligning with the Environmental Kuznets Curve (EKC) theory.Additionally, while trade openness may have short-term benefits, it ultimately harms the environment in South Africa in the long run.Factors such as FDI, industrial value-added, and energy use have negative effects on the environment, whereas technological innovations have been found to reduce CO 2 emissions, thus enhancing environmental quality.Moreover, the increasing population in South Africa is adding pressure to climate change and worsening the environmental condition.Adherence to the rule of law has been associated with lower CO 2 emissions, leading to improved environmental quality, while political stability and control of corruption have been linked to a deterioration in environmental quality, according to our findings.
Therefore, despite the necessity for economic growth, this study concluded that environmental sustainability should be South Africa's top priority.However, according to the findings, here are some recommendations: Global cooperation is vital to enhance environmental quality and transition to green economy.Strong international partnerships and comprehensive environmental provisions in trade agreements are necessary for achieving this goal.People should be made aware of the importance of adopting eco-friendly techniques in both production and consumption.For example, regardless of future environmental consequences, people still adhere to the traditional practice of utilizing whatever resources are available.Maintaining limits on carbon-intensive products and tapping into the potential of green economies can promote sustainable economic growth.This involves developing renewable energy, sustainable agriculture, recycling, and green financing.Lastly, it is worth noting that additional developmental policies could complement trade policy modifications to ensure that long-term advantages are obtained from reducing CO 2 emissions and to consistently encourage the advancement of new technologies that can improve domestic environment while also preserving the environment on a global scale.
However, considering the effects of trade on environmental degradation, future research could explore this more comprehensively, considering broader environmental indicators beyond CO 2 emissions such as methane and nitrous oxide.This could mean exploring how trade openness interacts with various environmental measures, thereby providing a more holistic view of the situation.
Technology and School of Management and Economics.We are extremely thankful to Professor Wei Yi-Ming for his endless support and guidance and to the editor and reviewers of this journal.This work was funded by Development Center of NFAGA, China, with the grant numbers 'JYC-2022-0046', 'JYC-2022-0049' and 'JYC-2023-0026'

Figure 2 .
Figure 2. The impulse response plot for scale effect (economic growth) and CO 2 emissions.Note.Based on the figure, it indicates a 10% change (increase and decrease) in economic growth and its effect on CO2 emissions.The dots on the figures signify average prediction value while the light blue and dark blue lines indicate 75%, 90% and 95% confidence interval respectively.

Figure 3 .
Figure 3.The impulse response plot for technical effect (economic growth squared) and CO 2 emissions.Note.Based on the figure, it indicates a 10% change (increase and decrease) in economic growth squared and its effect on CO2 emissions.The dots on the figures signify average prediction value while the light blue and dark blue lines indicate 75%, 90% and 95% confidence interval respectively.

Figure 4 .
Figure 4.The impulse response plot for trade openness and CO 2 emissions.Note.Based on the figure, it indicates a 10% change (increase and decrease) in trade openness and its effect on CO2 emissions.The dots on the figures signify average prediction value while the light blue and dark blue lines indicate 75%, 90% and 95% confidence interval respectively.

Figure 5 .
Figure 5.The impulse response plot for energy use and CO 2 emissions.Note.Based on the figure, it indicates a 10% change (increase and decrease) in energy consumption and its effect on CO2 emissions.The dots on the figures signify average prediction value while the light blue and dark blue lines indicate 75%, 90% and 95% confidence interval respectively.

Figure 6 .
Figure 6.The impulse response plot for foreign direct investment and CO 2 emissions.Note.Based on the figure, it indicates a 10% change (increase and decrease) in FDI and its effect on CO2 emissions.The dots on the figures signify average prediction value while the light blue and dark blue lines indicate 75%, 90% and 95% confidence interval respectively.

Figure 7 .
Figure 7.The impulse response plot for technological innovation and CO 2 emissions.Note.Based on the figure, it indicates a 10% change (increase and decrease) in technological innovation and its effect on CO2 emissions.The dots on the figures signify average prediction value while the light blue and dark blue lines indicate 75%, 90% and 95% confidence interval respectively.

Figure 8 .
Figure 8.The impulse response plot for industrial value added and CO 2 emissions.Note.Based on the figure, it indicates a 10% change (increase and decrease) in industrial value added and its effect on CO2 emissions.The dots on the figures signify average prediction value while the light blue and dark blue lines indicate 75%, 90% and 95% confidence interval respectively.

Table 1 .
Philips (2018)(2001)bles.Our study highlights that prior studies have employed the basic ARDL method built byPesaran et al (2001)and other co-integration frameworks.However, a new dynamic ARDL simulation model has been introduced by Jordan andPhilips (2018)to overcome the shortcomings of the basic ARDL technique.This innovative method has the capability to automatically simulate variable trends, generate forecast graphs, and estimate short and long-term relationships.Consequently, it enables swift prediction and plotting of probabilistic change forecasts on the dependent variable, based on a single explanatory variable and a fixed set of other regressors.The study used this method and its graphs to assess the impact of a specific explanatory variable and its actual rate of change on the dependent variable.

Table 3 .
Unit root test analysis.
Minlah and Zhang's (2021)he criterio.0.32% rise in CO 2 emissions, while a 1% increase in technical effect (measured by GDP square) results in a 0.54% reduction in CO 2 emissions.This implies that South Africa's economic growth has had a detrimental effect on its environment, while technical effects have contributed to its improvement, supporting the validity of the EKC hypothesis.These findings align with previous studies bySun et al (2021),Isik et al (2021), and Murshed (2021), but contradictMinlah and Zhang's (2021)assumption.

Table 5 .
ARDL bound test analysis.* and *** respectively represent statistical significance at 10%, 5%, and 1% levels.The respective significance levels suggest the rejection of the null hypothesis of no cointegration.The optimal lag length on each variable is chosen by the Schwarz's Bayesian information criterion (SBIC).