Asymmetric Effect of Natural Resource Exploitation on Climate Change in Resource-Rich African Countries

Abstract

This study investigated the asymmetric impact of natural resource exploitation on climate change in resource-rich African countries, based on panel data from 1980 to 2022. The dynamic common correlated effect (DCCE) and dynamic seemingly unrelated regression (DSUR) econometric techniques were employed to evaluate the long-term effects of positive shocks and negative shocks to natural resource exploitation. The findings revealed a positive relationship between both positive and negative shocks to natural resource exploitation and temperature, with increases in natural resource exploitation exerting a more intensified impact on temperature than decreases. In contrast, both positive and negative changes in natural resource exploitation are negatively related to precipitation, with an increased exploitation intensity having a more pronounced effect on rainfall patterns. The study also highlights the critical role of control variables such as GDP per capita, urban population, and total energy consumption in altering temperature and precipitation patterns. The findings underscore the importance of adopting sustainable natural resource extraction practices, integrating green technologies, and promoting collaboration across natural resource exploitation and renewable energy value chains to mitigate the negative impacts of natural resource exploitation.

1. Introduction

In recent years, climate change and sustainability have taken center stage as issues of urgent global concern. Undoubtedly, climate change is already widespread, and due to its significant risks to the environment, human welfare, and development, it is considered the greatest challenge of the 21st century (https://www.ucl.ac.uk/news/2009/may/climate-change-biggest-global-health-threat-21st-century, https://press.un.org/en/2021/sc14445.doc.htm—accessed on 12 August 2024). Proclamations have been made by scientific research bodies like the US National Assessments, the Intergovernmental Panel on Climate Change (IPCC), and the National Academy of Sciences that ‘global warming is unequivocal’ and that anthropogenic activities are the primary cause [1,2,3] (IPCC, 2012; NASEM, 2016; USGCRP, 2017). Human activity’s emission of greenhouse gases (GHGs) into the atmosphere is the primary cause of climate change, pushing the Earth’s temperature to a warmer level than it would naturally be (https://www.europarl.europa.eu/topics/en/article/20230316STO77629/climate-change-the-greenhouse-gases-causing-global-warming—accessed on 12 August 2024).

The Earth’s temperature has increased by more than one degree Celsius in recent decades, despite investors’ persistent initiatives to fund environmental quality and the green economy [4]. Unpredictable natural disasters, inevitable climatic changes, sea level rises, and meteorological events—such as intense rains in some parts of the planet and droughts, melting glaciers, and deserts in others—are other unfavorable environmental feedback that the Earth is currently grappling with [5] (IPCC, 2014). Forecasts suggest that climate change will persist throughout the remainder of this century and into subsequent decades (https://www.ipcc.ch/2021/08/09/ar6-wg1-20210809-pr/—accessed on 12 August 2024). Moreover, the IPCC projected that a 45% reduction in CO₂ emissions from 2010 levels by 2030 is required to keep global warming from surpassing 1.5 °C.

Studying the climate and global warming is crucial as it provides insight into the impact of anthropogenic activities on our planet’s environment and climate. Many anthropogenic activities produce huge volumes of GHGs into the atmosphere, trapping heat and raising our planet’s temperature. It has been argued that exposure to these GHGs may cause several negative health outcomes such as respiratory issues, cancer, heart disease, mental impairment, and even death [6]. Moreover, these activities likely engender long-term repercussions such as rising sea levels, a decline in biodiversity, and an increase in the frequency and intensity of extreme weather events. According to [7], the temperature change will cause hitherto unheard-of repercussions, such as ocean acidification, rising sea levels, warmer marine temperatures, and melting glaciers, in addition to increasing ocean temperatures due to extra heat and energy. These result in elevated levels of toxicity in water resources, causing pollution and the demise of aquatic biodiversity [8]. By implication, coastal communities’ means of subsistence are severely impacted, including the almost 60 million people employed in the fisheries and aquaculture industry globally [9], the 40% of the global population who reside 100 km or less from the coast (https://www.un.org/esa/sustdev/natlinfo/indicators/methodology_sheets/oceans_seas_coasts/pop_coastal_areas.pdf—accessed on 12 August 2024), and the 898 million people who live on low-lying shores [10].

This study explores the asymmetric effect of natural resource (NR) exploitation on climate change in resource-rich African countries. Among anthropogenic activities, NR exploitation provokes debate about the paradox of progress and devastation. Ref. [11] demonstrated that mining mineral resources can concurrently advance certain economic Sustainable Development Goals (SDGs) while harming other social and environmental SDGs. Moreover, scholars hold the perspective that despite the large economic gains, the extraction of NRs often brings about significant negative social impacts, with the local population near the extraction sites predominantly affected. Several types of NRs, especially minerals, are prominent for their huge lifecycle-long impact (from exploration and extraction to utilization and transportation) on pollution and waste generation [12,13]. The mining industry is a substantial source of GHG emissions, as the extraction of minerals often involves energy-intensive processes that rely on fossil fuels, leading to significant carbon emissions. For example, mining operations and the transportation of ores and processed minerals contribute to direct emissions, such that as of 2021, their contribution to global emission was estimated at 4–7% (https://www.weforum.org/agenda/2021/09/clean-energy-reliant-on-sustainable-mining-and-cement/—accessed on 12 August 2024), while ref. [14] found that every year, global mining and resource extraction activities contribute to GHG emissions leading to up to €5 trillion of damage.

The processing of NRs, such as smelting and refining, is highly energy intensive. The energy required for these processes is often derived from fossil fuels, leading to further carbon emissions. Ref. [15] pointed out that the energy consumption to produce NRs contributes significantly to GHG emissions, accounting for a considerable share of the industrial sector’s total emissions. The development of NR extraction sites frequently necessitates deforestation and the significant alteration of land, which reduces the carbon sequestration capacity. Forests act as carbon sinks, and their removal releases stored carbon into the atmosphere. Ref. [16] emphasizes that the deforestation and land use changes associated with mining operations result in the release of stored carbon and a decrease in the land’s ability to absorb CO₂, thereby contributing to climate change. Coal mining is a notable contributor to methane (CH₄) emissions, a potent greenhouse gas. Methane is released during both coal extraction and from disused mines. As highlighted by [5], methane exhibits a global warming potential (GWP) that is 28–36 times higher than CO₂ over a 100-year period, underscoring its significant impact on climate change dynamics when emitted during coal mining operations.

The United Nations (UN) has established the Sustainable Development Goals (SDGs) to mitigate environmental challenges. These SDGs serve as a roadmap for nations to achieve sustainability, yet many countries struggle with dilemmas surrounding industrial production and the urgent action needed to combat climate change, in line with SDG 13. Moving toward renewable and sustainable energy consumption is integral to the attainment of SDG 13 (https://sdgs.un.org/goals/goal13—accessed on 12 August 2024), and NRs have been identified as a crucial requirement for this transition. It has been argued that exploiting NRs for renewable energy extraction is becoming increasingly necessary as we move toward clean energy [17]. Minerals like silicon, cadmium, and tellurium are crucial to produce solar panels. Rare earth elements such as neodymium and dysprosium are essential for manufacturing strong magnets utilized in wind turbines. The increased availability of these minerals can support the expansion of wind and solar energy infrastructure and reduce the reliance on fossil fuels and NR depletion (https://ecampusontario.pressbooks.pub/environmentalscience/chapter/chapter-12-resources-and-sustainable-development/—accessed on 12 August 2024). Minerals such as lithium, copper, nickel, and aluminum are vital for manufacturing energy-efficient electrical components and infrastructure [18].

Minerals like gypsum and limestone are used to produce energy-efficient building materials, such as drywall and cement. These materials can improve insulation and reduce energy consumption for heating and cooling in buildings. It has been noted that the rapid adoption of electric vehicles faces hurdles due to the inadequate supply of essential minerals [19]. Furthermore, minerals like basalt and peridotite are essential for climate mitigation technologies like mineral carbonation and carbon capture [20], while a circular economy is facilitated as the efficient recycling of metals like aluminum and copper can reduce the need for new mining, lowering energy consumption and minimizing environmental impacts [21]. According to [22], the increasing mineral demand, fueled by the production of renewable energy, underscores the need to optimize NRs to achieve the goal of a decarbonized future before the 22nd century. Therefore, it is unsurprising that the minerals market for energy transition has grown twice as large over the past five years, achieving $320 billion in 2022 (https://mine.nridigital.com/mine_oct23/brics-critical-mineral-trade-africa—accessed on 12 August 2024).

Arguments have also emerged in the literature that the implementation of cutting-edge green technology in the exploitation of NRs could be an effective way of mitigating the environmental impact of utilizing NRs [23,24]. Advancements in mining technologies, such as precision mining and automation, can reduce the environmental footprint of mining activities, including GHG emissions (https://www.ief.org/news/how-to-make-mining-more-sustainable—accessed on 12 August 2024). Also, the use of biomining, liquid emulsion technologies, dust-suppression techniques, and sustainable mining practices like the 3Rs technique (reduce, recycle, and reuse) and decarbonizing through advanced carbon capture and storage technologies have proven effective in mitigating the negative externalities from mineral exploitation [25]. With these technologies, more NRs can be explored with a minimal impact on climate change. This argument aligns with a World Bank report that posited that NRs are crucial for advancing a green, low-carbon future, central to achieving the Paris Agreement, and essential for the effective execution of green recovery plans after COVID-19, which also stressed the importance of stakeholders in both minerals and the renewable energy value chain collaborating toward meeting the SDG 7, sans negative impacts on the climate [26].

Despite the paradoxical consequences of NR exploitation, research on the asymmetric impact of NR exploitation, exploring the effect of positive shocks and negative shocks on NR exploitation, has been sparse. Particularly, studying the asymmetric impact of NR exploitation on climate change in Africa is crucial given the unique position these countries hold in the global economy and their immense contribution to both GHG emissions and NR production. Many African countries rely heavily on the extraction and export of natural resources, particularly oil, gas, minerals, and timber. The exploitation of these resources contributes significantly to GHG emissions, which is a major driver of climate change. Conversely, African countries are often highly susceptible to climate change’s effects, such as severe weather, droughts, floods, and changes in agricultural productivity (https://wmo.int/media/news/africa-suffers-disproportionately-from-climate-change—accessed on 12 August 2024). Furthermore, the extraction of natural resources can lead to significant social and economic challenges, including inequality, conflict, and poor governance. These challenges further complicate the efforts to mitigate or adapt to climate change. For example, it has been established that corruption and the mismanagement of resource wealth exacerbate inequality in African countries [27], hindering investment in climate resilience.

Therefore, this study addresses a gap in the existing literature by investigating the asymmetric effect of NR exploitation on climate change, specifically within the context of resource-rich African countries. Previous studies have largely focused on either the emission consequences of resource extraction or the environmental impacts of specific activities without examining the nuanced effects of positive and negative shocks to NR exploitation on climate change, especially in the context of Africa. Though the study by [28] considered asymmetric effects in the context of African countries, the study’s focus was carbon emissions and not climate change. Africa’s dual role as a significant contributor to global NR production and a region disproportionately affected by climate change¹¹ amplifies the urgency of this investigation. Furthermore, the interplay between resource governance challenges, such as corruption and inequality, and climate resilience remains underexplored. This study seeks to fill these gaps by providing an in-depth analysis of how the asymmetric impacts of NR exploitation manifest in resource-rich African nations, offering insights critical for sustainable NR management and climate policy in the region. As such, understanding the asymmetric impacts of NR exploitation on climate change within these countries can illuminate both the costs and opportunities of resource management in diverse socio-economic and environmental contexts.

The remainder of the study is structured as follows. Section 2 consists of a review of the relevant literature, while Section 3 discusses the methodology. Section 4 contains a discussion of the study’s results, and Section 5 concludes the study.

2. Literature Review

Many research studies have investigated how NRs, environmental conditions, and climate change are interconnected across various regions and countries around the world [28,29,30,31,32,33,34,35]. The discussions and findings of scholars regarding the impact of NR extraction and consumption have affected the debate surrounding the paradox of development versus destruction. The mining and consumption of NRs involve several operations and activities, including excavation, drilling, construction, refining, mine reopening, transportation, waste management, usage, and disposal, with each impacting the environment in various ways. As such, many studies have emphasized the negative impact of NR exploitation on climate change. On the other hand, some scholars have argued that when managed responsibly, NR exploitation can support environmental sustainability by enabling technologies and practices that reduce pollution, conserve resources, and promote ecosystem health.

According to [4], the exploitation of NRs plays a significant role in exacerbating global warming, which, in turn, can indirectly contribute to the intensification of climate change. The extraction, processing, and transportation of NRs can engender the release of GHGs that contribute to global warming, increase sea levels, and accelerate the melting of glaciers and icecaps [36]. Moreover, ref. [37] argued that certain mineral extraction practices can intensify the impacts of sea level rise in coastal areas, while also causing land degradation and disrupting the natural flow of water resources. To examine the effect of NR exploitation on the green growth of Central Asian countries, ref. [32] analyzed data based on the ARDL method and found that the rise in the extraction of NRs negatively affects the region’s green growth index, driven by detrimental environmental effects.

A study by [38] on South Africa revealed that the country’s coal production has a significant potential for global warming, contributing up to 95%. This supports [39] who argued that the rise in global warming in South Africa can largely be attributed to the growing reliance on coal for energy. A greater concern is the potential health risks posed by the harmful gases released from coal mines. Exposure to these gases is believed to lead to a range of health problems, such as respiratory issues, cancer, cardiovascular diseases, mental impairments, and even death [40]. The impact of mining coal minerals on surface water resources was investigated by [41] for Australia by evaluating 110 surface water parameters. The test results of the study detected higher metal levels in water resources near the coal mines than those in other areas.

Moreover, ref. [42] explored the impact of mining on the environmental quality of India’s Talcher coalfield and found that besides worsening air pollution in the area due to the increased use of unsafe coal transportation, heavy vehicles, and the lack of a water-spraying system, the local water bodies were also negatively affected, with some even dropping below acceptable standards. In the case of China, ref. [43] posited that the country’s economic growth is fueled by the rapid development of its NR and that this expansion has significantly contributed to environmental pollution and global warming. Moreover, in a study on the linkage between NR, urbanization, economic growth, human capital, and the ecological footprint, ref. [44] applied the ARDL, FMOLS, DOLS, and CCR techniques to analyze data from South Africa and found that the utilization of NRs, alongside urbanization and economic growth, increased the ecological footprint. These results support [45] in their analysis of the effect of NR rent and economic growth on carbon emissions. Using the panel mean group technique to estimate data from 16 EU countries, the study found that NR rent, economic growth, and energy consumption exert a long-term negative impact on environmental quality. Also, ref. [46] investigated data from Malaysia using the autoregressive integrated moving-average method and found that NR heightened carbon emissions.

Furthermore, by estimating a cross-sectional ARDL model, ref. [47] explored the linkage between NR investment and carbon emissions in China. The estimates confirmed that NR is associated with an increase in carbon emissions. In a similar study in China, ref. [48] researched the determinants of carbon emissions and found that China’s exploitation of NRs heightened CO₂ emissions. In a similar study conducted on the G7 countries, Wang et al. (2020) investigated how financial development and NRs influence the regulation of CO₂ emissions. The study’s empirical results indicated that both economic growth and NRs contribute to an escalation in CO₂ emissions.

Another area of the literature underscores the transition to clean energy facilitated by mineral resources, indicating a positive relationship between NRs and climate change due to their supportive role in advancing green energy. In practice, an ample supply of NRs is vital for maintaining energy security and enabling the energy transition, with industrialized countries facing risks from potential resource shortages [49]. In this context, the expansion of green technologies is linked to increased availability of mineral resources, demonstrating that the positive relationship strengthens over time [50,51]. For this to be sustained, it is crucial to ensure an adequate supply of minerals such as graphite, lithium, cobalt, and rare earth elements like dysprosium, terbium, praseodymium, and neodymium, as these materials are essential for the transition to sustainable energy sources. Although this objective is crucial, there are obstacles to the broad use of electric vehicles due to the limited supply of these essential minerals [19]. Replacing these components is difficult as it either threatens the availability of other scarce minerals or requires major changes to manufacturing processes overall. According to [52], the shortage of material resources is a significant challenge in preserving energy security and reducing carbon emissions.

The increasing mineral demand, fueled by the production of renewable energy, highlights the critical need to optimize NRs to achieve the end-of-the-21st-century deadline for delivering a decarbonized future. For example, by exploring the role of green development in fostering sustainable growth in Kyrgyzstan, ref. [53] highlighted solar and wind energy resources as key potential development opportunities, dependent on sufficient investment and technology transfer from more developed economies. The challenge of climate change in Central Asia was the focus of the study by [54], highlighting the need for green economic sectors and a reduction in reliance on non-renewable energy resources as key priorities for mitigating climate change risks. In a similar study, ref. [55] highlighted the importance of NRs as a key driver of economic growth and industrial production, while also emphasizing the need to improve NR efficiency and lower NR intensity through process reforms and technological innovations to reduce their adverse effects. Moreover, ref. [56] stressed the importance of the efficient utilization of scarce mineral resources, arguing that the sustainable use of geologically limited minerals is essential for achieving environmentally sustainable growth. Furthermore, in addressing the need to lessen the dependence on mineral resources for green growth, ref. [57] concluded that enhancing mineral resource efficiency is a global priority focused on environmental protection and the renewal of NRs.

Several studies have underscored the relationship between energy productivity, driven by the use of metals, minerals, and energy, and the efforts to mitigate climate change. A study by [58] investigated how electricity, NRs, and economic growth influenced carbon emissions in five EU economies. The estimates of their panel least-squares model demonstrated that NRs and renewable electricity helped reduce CO₂ emissions. In a similar study, ref. [59] found that NRs could be leveraged to control CO₂ emissions in the USA. In the study by [45], NR rent was reported to mitigate CO₂ emissions in the short term, while it intensified it in the long term. Mixed results were similarly reported by [60] after employing the AMG model to estimate data from BRICS. Though the AMG estimates showed that the relationship between economic development and CO₂ emissions supported the Environmental Kuznets Curve (EKC) hypothesis in these economies, the study’s findings revealed that NR extraction helped to reduce pollution in Russia, whereas they contributed to CO₂ emissions in South Africa. In another study on BRICS, ref. [61] utilized DOLS and FMOLS methods to assess the EKC hypothesis. The researchers found that the extraction of NRs, the use of green energy, and urbanization contributed to a reduction in the ecological footprint of BRICS countries.

The EKC in Tunisia was evaluated by [62]. The results of the canonical cointegrating regression conducted in the study validated the presence of EKC in Tunisia. The authors further contended that the extraction of NRs intensified carbon pollution in the country. Furthermore, ref. [63] found an inverted U-shaped relationship between wealth and carbon pollution in Africa. In contrast, he also established a U-shaped link between income and ecological footprints in the same region. Further research by [64] in Africa investigated the relationship between wealth growth and carbon emissions in 49 countries using certain panel estimation techniques including a semi-parametric fixed-effects method. Their findings supported a nonmonotonic and upward-trending relationship between the two variables while disproving the EKC hypothesis. Moreover, upon investigating the development–carbon emissions nexus in various African regions, refs. [65,66] both found a U-shaped relationship rather than the classic EKC for the Economic Community of West African States (ECOWAS) and East African countries.

The presence of asymmetry in the relationship between NR utilization and environmental outcomes has also been given attention by researchers recently. Ref. [28] examined the reaction of carbon emissions in Africa to both symmetric and asymmetric shocks from resource consumption, economic growth, urbanization, and energy use. By estimating a panel ARDL-PMG model for 44 African economies, it was found that both positive and negative shocks to NR consumption had a significant positive impact on carbon emissions, in contrast to the insignificant effect observed for the symmetric estimation. In a similar study, ref. [67] focused on China and investigated the linkage between NR, human capital, and sustainable development. The results of their econometric analysis highlight an asymmetric effect of mineral and oil rents on economic growth, with forest rents exhibiting a positive and significant impact. Furthermore, the factors influencing carbon emissions in emerging economies in the past thirty years were investigated by [68]. The study’s findings show that NR factors have asymmetric effects on carbon emissions, with oil rents lowering emissions and mineral rents leading to an increase. Similarly, ref. [69] explored the symmetric and asymmetric effects of NR utilization on Saudi Arabia’s ecological footprint between 1981 and 2018, using the dynamic ARDL estimation technique. The symmetric analysis revealed that oil rent, natural gas rent, and total rent are positively linked to the ecological footprint, while mineral resources showed no significant impact. On the other hand, the asymmetric analysis showed that positive shocks to oil rent, natural gas rent, and total rent heighten the ecological footprint, while the decrease in these NR rents had no impact.

3. Methodology

3.1. Model Specification

To investigate the asymmetric relationship between NR exploitation and climate change, this study drew on the Stochastic Impacts by Regression on Population, Affluence, and Technology (STIRPAT) model [70,71], which can be expressed as follows:

$$I_{it}={\alpha P}_{it}^b{A}_{it}^c{T}_{it}^d{\epsilon }_{it}$$

(1)

where I, P, A, and T denote the environmental impact, population, affluence, and technology, respectively. $\alpha$, $i$, and $t$ represent the constant term, individual countries, and time, respectively, while $\epsilon$ represents the error term.

By taking the logarithms, Equation (1) becomes:

$$\ln {(I}_{it})=\alpha +b\ln P_{it}+c\ln A_{it}+d\ln T_{it}+{\epsilon }_{it}$$

(2)

In this study, I denotes climate change (CC) as a proxy for environmental degeneration, while P, A, and T represent urban population, real GDP per capita as a proxy for economic growth, and total energy consumption, respectively. To account for the impact of NR exploitation, Equation (2) is augmented by total NR rents to produce the following model:

$$\ln {CC}_{it}={\alpha }_0+{\alpha }_1\ln {NR}_{it}+{\alpha }_2\ln {URB}_{it}+{\alpha }_3\ln {GDP}_{it}+\ln {EC}_{it}+{\epsilon }_{it}$$

(3)

Following extant research [69,72,73,74], NRs in Equation (3) are decomposed into their positive and negative components such that:

$${NR}_{it}^{+}=\sum _{j=1}^k{∆NR}_j^{+}=\sum _{j=1}^k\mathrm{m}\mathrm{a}\mathrm{x}(∆{NR}_{ij},0)$$

(4)

and

$${NR}_{it}^{-}=\sum _{j=1}^k{∆NR}_j^{-}=\sum _{j=1}^k\mathrm{m}\mathrm{i}\mathrm{n}(∆{NR}_{ij},0)$$

(5)

By substituting Equations (4) and (5) into (3), Equation (3) becomes:

$$\ln {CC}_{it}={\alpha }_{0i}+{\alpha }_{1i}{\mathrm{l}\mathrm{n}NR}_{it}^{+}+{\alpha }_{2i}{\mathrm{l}\mathrm{n}NR}_{it}^{-}+{\alpha }_{3i}\ln {URB}_{it}+{\alpha }_{4i}\ln {GDP}_{it}+{\alpha }_{5i}\ln {EC}_{it}+{\epsilon }_{it}$$

(6)

In Equation (6), the asymmetry hypothesis centers on the estimations of ${\alpha }_{1i}$ and ${\alpha }_{2i}$, as the impact of NRs on CC is considered asymmetric if and only if both parameters are statistically significant and exhibit varying magnitudes. If so, it is determined that both have different impacts on climate change in terms of their nature and/or intensity. However, the effect of NRs on climate change is considered linear rather than asymmetric if one or both requirements are broken.

3.2. Estimation Procedure

The estimation of data was carried out in four phases. The first phase involved investigating whether cross-sectional dependence (CSD) exists in the data. The [75] in-series CSD test was implemented to evaluate whether the panel data were characterized by CSD. In the second phase, the integration properties of the variables were assessed by conducting second-generation panel unit root tests, which have the capacity to accommodate the problem of CSD in panel data. The implemented tests were the [76] cross-sectional augmented ADF (CADF) and cross-sectional augmented IPS (CIPS). The CADF test statistic is stated as follows:

$${∆Z}_{it}={\rho }_i+a_i{z}_{i,t-1}+{\beta }_i{\overline{z}}_{t-1}+\sum _{j=0}^k{b}_{ij}{∆\overline{z}}_{t-j}\sum _{j=1}^k{\gamma }_{ij}{∆z}_{i,t-j}+{\epsilon }_{it}$$

(7)

where the lagged level of cross-sectional averages is denoted by ${\overline{z}}_{t-1}$, while the first-order integration of each cross-section is represented by ${∆\overline{z}}_{t-j}$.

The CADF generates the CIPS test statistics and is presented as follows:

$$\mathrm{CIPS}=N^{(-1)}{\sum }_{(i=1)}^{N}CADFi$$

(8)

The third phase of the estimation process involved testing cointegration among the variables. The [77] ECM-based cointegration test was utilized, considering its capability to accommodate the presence of CSD in panel data, handle heterogeneous panels, and produce unbiased results. This test computes four test statistics in two categories: group statistics (Gt and Ga test statistics) and panel statistics (Pt and Pa test statistics). To estimate the [77] test, the following least-squares model is used:

$${∆Z}_{it}={\rho }_i^{\prime }{d}_t+{\delta }_i\left(Z_{it-1}-{\alpha }_i\prime x_{it-1}\right)+\sum _{j=1}^{pi}{\delta }_{ij}{∆z}_{it-j}+\sum _{j=-pi}^{pi}{\theta }_{ij}{∆x}_{it-j}+{\epsilon }_{it}$$

(9)

The last phase of the econometric procedure entails estimating long-run parameters to determine the asymmetric impact of NR exploitation on climate change. This study employed the dynamic common correlated effect (DCCE), which is notable for addressing several key panel data problems that other traditional estimation approaches cannot. First, the DCCE method accommodates CSD in panel data by obtaining the average and lags of every cross-sectional unit [78]. Second, the parameter heterogeneity problem is handled by the DCCE through the deployment of its mean group component. Moreover, the DCCE method eliminates asymptotic distortion caused by the endogeneity of the independent variables and accounts for non-stationarity in the data [79]. The DCCE handles endogeneity in both static and dynamic panel data models adequately by creating an instrument set using the lagged form of the variables. Particularly in dynamic models, DCCE enables regressors to greatly enhance the estimator’s small sample properties whether they are endogenous, weakly exogenous, or strictly exogenous [80]. Following [79], the DCCE model can be stated as follows:

$$\ln {CC}_{it}={\alpha }_i\ln {CC}_{it-1}+{\delta }_i{X}_{it}+\sum _{p=0}^{P_T}{\gamma }_{xip}{\overline{X}}_{t-p}+\sum _{p=0}^{P_T}{\gamma }_{yip}{\overline{X}}_{t-p}+{\mu }_{it}$$

(10)

where lnCC denotes the natural logarithm of climate change whereby its lag is an independent variable, the grouping of the remaining independent variables is denoted by $X_{it}$, and the lagged average of cross sections is denoted by $P_T$.

The robustness of the DCCE estimates was assessed by estimating a dynamic seemingly unrelated regression (DSUR) introduced by [81]. This technique expands on the single-equation DOLS by applying it to panel datasets in which there are more time dimensions than cross-sections. According to [81], the DSUR approach is effective in handling CSD, heterogeneity, and endogeneity in panel data.

Lastly, causal relationships among the variables of interest were evaluated by using the [82] panel causality test, reputed for its ability to handle the issues of CSD and heterogeneity in panel data.

3.3. Data

The annual data of 10 resource-rich African countries were estimated in this study over the period of 1980 to 2022. The study period was selected based on the availability of data. The countries involved in the study were Algeria, Angola, the Democratic Republic of Congo, Egypt, Gabon, Ghana, Libya, Nigeria, South Africa, and Zimbabwe. The dependent variable is climate change, and it was measured by two variables: average temperature and precipitation, in consistency with prior research [83,84] and provided by the World Bank’s Climate Change Knowledge Portal. The main explanatory variable is NR exploitation, measured by the total natural resource rent as a percentage of GDP, in line with previous research [44,85]. The control variables consist of urbanization, economic growth, and energy consumption, measured by urban population (% of the total population), GDP per capita, and energy use (kg of oil equivalent per capita). Alongside the data for the total natural resource rent, control variables’ data were from the World Bank’s World Development Indicator.

4. Discussion of Results

The descriptive statistics for the variables are shown in

Table 1. Descriptive statistics.

Variable	Mean	Std. Dev.	Minimum	Maximum
Temperature	23.477	2.720	17.100	28.070
Precipitation	825.945	618.694	11.210	2233.360
Natural resources rent	18.843	13.041	2.489	66.059
Urban population	48.555	16.629	21.970	87.651
GDP per capita	3595.173	3165.213	322.440	13,729.16
Energy use	1086.977	865.211	257.781	3243.823

. The average temperature for the countries under study between 1980 and 2022 was 23.48 °C. The highest average temperature of 28.07 °C was recorded in 2010 by Ghana, while the lowest average temperature of 17.10 °C was recorded in 1981 by South Africa. The average precipitation was 825.95 mm, with the highest precipitation recorded being 2233.36 mm in 2007 by Gabon, and the lowest precipitation being 11.21 mm in 2010 recorded in Egypt. The average NR rent over the period was 18.84% of GDP. Libya recorded the highest NR rent/GDP ratio of 66.06 in 2008, while the lowest of 2.49 was recorded by South Africa in 1999.

The estimation procedure started with an in-series CSD test using the technique of [75], the results of which are presented in

Table 2. Cross-sectional dependence test.

	TEMP	PREC	NR+	NR−	URB	GDP	EC
CSD test	10.355 ***	8.349 **	7.132 ***	15.017 ***	11.957 **	18.834 ***	8.446 ***
p-value	0.000	0.031	0.000	0.000	0.010	0.000	0.000

Note: All variables are in logarithm forms; *** and ** indicate 1% and 5% levels of significance, respectively; TEMP = Average temperature, PREC = Precipitation, NR+ = Positive shock to NR exploitation, NR− = Negative shock to NR exploitation, URB = urban population, GDP = GDP per capita, EC = Total energy use.

. All the tested variables returned significant probability values, which suggests the rejection of the null hypothesis that there is no CSD in the variables. These results imply that all the variables are characterized by CSD and necessitate the use of unit root tests that accommodate CSD in panel data to ensure reliable estimates. To this end, this study applied the [76] second-generation unit root tests, namely CADF and CIPS, given their ability to mitigate CSD in panel data. As displayed in

Table 3. Panel unit root tests.

Variable	CADF		CIPS
	Level	First Difference	Level	First Difference
TEMP	−1.527	−5.307 ***	−0.992	−8.621 ***
PREC	−2.113	−5.437 **	−1.351	−5.424 ***
NR+	0.894	−6.638 ***	0.944	−7.809 **
NR−	1.117	−4.201 **	−1.521	−2.331 **
URB	2.013	−3.384 ***	−2.007	−4.794 ***
GDP	−1.841	−6.441 ***	1.346	−9.162 **
EC	0.982	−3.865 ***	−1.230	−3.550 ***

Note: All variables are in logarithm forms; *** and ** indicate 1% and 5% levels of significance, respectively; TEMP = Average temperature, PREC = Precipitation, NR+ = Positive shock to NR exploitation, NR− = Negative shock to NR exploitation, URB = urban population, GDP = GDP per capita, EC = Total energy use.

, both CADF and CIPS tests confirm that all the variables are not stationary at level, but all become stationary after the first difference, confirming their levels of integration as I(1) processes. These results imply the need to evaluate the long-term relationship between the variables through a test of cointegration.

The Westerlund ECM-based test was conducted to assess the cointegration among the variables, and

Table 4. Panel cointegration test results.

Statistic	No Constant		Constant		Constant and Trend
	Value	p-Value	Value	p-Value	Value	p-Value
Gt	−2.217 ***	0.000	−5.822 ***	0.000	−1.649 ***	0.000
Ga	1.006 **	0.010	−3.219 ***	0.000	0.893 **	0.015
Pt	−2.382 ***	0.000	−6.384 **	0.028	−2.649 **	0.030
Pa	−0.981 ***	0.000	−2.571 ***	0.000	2.311 ***	0.000

Note: All variables are in logarithm forms; *** and ** indicate 1% and 5% levels of significance, respectively.

presents the results. The test was conducted in three forms: without a constant; with a constant; and with a constant and trend. The decision for establishing whether the variables are cointegrated involves testing the null hypothesis of no cointegration among the variables based on the test’s group and panel statistics. As reported in the table, all four statistics returned significant probability values in all three forms of the test, indicating that the null hypothesis was rejected for all four statistics. These results confirm that all the variables in the study are cointegrated.

The long-term impacts were assessed based on the DCCE approach. To check whether the results were robust, the DSUR technique was subsequently applied.

Table 5. DCCE results.

Variables	Model 1: Dependent Variable = Temperature		Model 2: Dependent Variable = Precipitation
	Coefficient	p-Value	Coefficient	p-Value
NR+	0.118 ***	0.000	−0.207 ***	0.000
NR−	0.047 ***	0.000	−0.051 **	0.036
URB	−0.094 **	0.015	0.113 **	0.018
GDP	0.186 ***	0.000	−0.226 ***	0.004
EC	0.231 ***	0.000	−0.094 **	0.027

Note: All variables are in logarithm forms; *** and ** indicate 1% and 5% levels of significance, respectively; NR+ = Positive shock to NR exploitation, NR− = Negative shock to NR exploitation, URB = urban population, GDP = GDP per capita, EC = Total energy use.

and

Table 6. DSUR results.

Variables	Model 1: Dependent Variable = Temperature		Model 2: Dependent Variable = Precipitation
	Coefficient	p-Value	Coefficient	p-Value
NR+	0.064 **	0.022	−0.168 ***	0.000
NR−	0.012 ***	0.000	−0.045 ***	0.006
URB	0.121 **	0.039	0.064 **	0.011
GDP	0.246 ***	0.000	0.107 ***	0.000
EC	0.139 ***	0.000	0.083 **	0.043

Note: All variables are in logarithm forms; *** and ** indicate 1% and 5% levels of significance, respectively; NR+ = Positive shock to NR exploitation, NR− = Negative shock to NR exploitation, URB = urban population, GDP = GDP per capita, EC = Total energy use.

present the results of the DCCE and DSUR long-term estimations, respectively. In each regression, two different regressions were carried out. The first regression utilized temperature, which measures climate change, as the dependent variable, and the results are presented in the two tables as Model 1. The second regression utilized precipitation, which also measures climate change, as the dependent variable, and the results are presented in the two tables as Model 2. Comparing the results from both DCCE and DSUR demonstrates that the results are mostly consistent across both estimation methods. As a result, similar interpretations and discussions are applicable to the results from both sets of findings.

To test the asymmetric relationship between NR exploitation and climate change, NR rent was decomposed into positive and negative components. The DCCE results in Model 1 of Table 5 show that both positive and negative changes in NR rents are positive and statistically significant. This implies that positive and negative shocks to NR rents have a positive relationship with temperature, implying that both increases and decreases in NR exploitation are associated with increases in temperature in the long term. However, considering their respective coefficients, an increase in NR exploitation was found to impact temperature with more intensity compared to a decrease in NR exploitation, as the coefficient of NR+ is greater than that of NR−. These findings corroborate [86] who argued that environmental degradation is significantly engendered by the consumption of NR through mining activities, deforestation, and agricultural practices. For example, resource-rich African countries such as the DRC and Gabon rely heavily on logging and mining, leading to large-scale deforestation. Deforestation reduces the carbon sequestration potential of forests, increasing atmospheric CO₂ levels and contributing to global warming. In the DRC, mining activities for minerals like cobalt and copper result in habitat destruction, accelerating localized warming..

Many African nations primarily rely on exploiting their natural resource wealth and engaging in agricultural activities to drive economic growth. These sectors play a critical role in shaping the economies of the continent, with NR extraction and farming being central to income generation, employment, and overall development. Due to weak environmental regulations and widespread corruption in many African countries [87], multinational companies (MNCs) operating in the extractive sector often fail to adopt environmentally sustainable practices. The lack of effective enforcement and accountability mechanisms allows these companies to prioritize profit over environmental protection, leading to significant ecological degradation. For example, Nigeria’s Niger Delta experiences significant CO₂ and CH₄ emissions due to gas flaring by MNCs, which contribute to regional warming trends [88,89]. Additionally, previous research found that the coal mining process in South Africa has a massive potential of 95% to contribute to global warming [38], and coal prevalence as a source of energy was largely behind the country’s increased temperature [39].

The DCCE asymmetric results in Model 2 of Table 5 show that both positive and negative changes in NR rents are negative and statistically significant. This implies that positive and negative shocks to NR rents have a negative relationship with precipitation, implying that both increases and decreases in NR exploitation are associated with a decrease in precipitation in the long run. However, considering their respective coefficients, an increase in NR exploitation is found to impact precipitation with more intensity compared to a decrease in NR exploitation, as the coefficient of NR+ is greater than that of NR−. These findings could be attributed to the nature of NR exploitation in resource-rich African countries. The intensification of NR exploitation (such as logging, mining, oil extraction, and industrial agriculture) in these countries could impact regional and global precipitation patterns through deforestation, GHG emissions, land degradation, and changes in atmospheric circulation. For example, forests are very crucial to the hydrological cycle, as they recycle water through evapotranspiration, which contributes to cloud formation and precipitation. Deforestation in countries like DRC and Gabon, driven by mining and logging activities, disrupts these processes, leading to decreased rainfall and more prolonged dry periods. It has been found that the Congo Basin, home to one of the world’s largest tropical rainforests, influences regional precipitation through moisture recycling and that deforestation in this basin could result in a 10–20% decrease in rainfall in the surrounding areas [90]. In Algeria, the overexploitation of aquifers has been found to exacerbate desertification, reducing local precipitation and increasing dependency on irregular rainfall [91].

Moreover, NR exploitation often emits huge amounts of GHGs, which affects precipitation patterns such as the African monsoon and the Intertropical Convergence Zone (ITCZ), altering rainfall distribution. It has been found that resource-rich African countries like Algeria, Egypt, and Libya experienced changes in rainfall patterns due to shifts in the ITCZ caused by increased desertification and GHG emissions, which engendered reduced precipitation in arid and semi-arid zones [92]. In Angola and Nigeria, offshore oil extraction can affect coastal ecosystems, such as mangroves and seagrasses, which play roles in regulating local climate and rainfall. The degradation of these ecosystems weakens their ability to influence precipitation patterns. Specifically, Angola’s coastal regions have seen changes in precipitation due to the destruction of mangroves and marine habitats, which disrupt local hydrological cycles [93].

Now turning to the control variables, the urban population is found to exert a negative impact on temperature and a positive impact on precipitation. This result corroborates a previous study in Africa by [85] who argued that urban dwellers have greater access to higher-education opportunities, which in turn enhances their environmental awareness and encourages them to take proactive measures in caring for the environment. The results also show that GDP per capita increases temperature and reduces precipitation as shown in Model 1 and Model 2, respectively. Economic growth in most African nations, and across the continent in general, is primarily fueled by the agricultural sector. Agriculture is a key driver, contributing significantly to employment, GDP, and export earnings, thereby supporting broader economic development. As discussed earlier, agricultural activities involve deforestation, which increases temperature and undermines rainfall patterns. This result is consistent with the previous results of [85], who found that economic growth is deleterious to environmental outcomes in Africa. However, this result contradicts [30] who found that economic growth facilitates green growth in the OECD. The coefficients of energy consumption show that the variable increases temperature (Model 1) and reduces precipitation (Model 2). This implies that the intensified consumption of energy, especially in the process of NR exploitation, exerts a deleterious impact on climate change. This research outcome is consistent with the finding of [28] that energy consumption intensifies carbon emissions in Africa.

Determining the causal relationships between the key variables is the final step in this study’s econometric process. For resource-rich African countries aiming to mitigate the climate change impact of NR exploitation, this test will provide guidance on which variables to prioritize for optimal impact.

Table 7. Dumitrescu Hurlin panel causality test.

Null Hypothesis.	W-Statistic	Zbar-Statistic	p-Value
PREC does not homogeneously cause TEMP PREC does not homogeneously cause TEMP	3.032 4.175 ***	1.286 2.896	0.198 0.004
NR does not homogeneously cause TEMP TEMP does not homogeneously cause NR	3.396 * 1.714	1.774 −0.573	0.076 0.567
URB does not homogeneously cause TEMP TEMP does not homogeneously cause URB	10.517 *** 3.042	11.824 1.305	0.000 0.192
GDP does not homogeneously cause TEMP TEMP does not homogeneously cause GDP	8.370 *** 3.896 **	8.801 2.508	0.000 0.012
EC does not homogeneously cause TEMP TEMP does not homogeneously cause TEMP	5.366 *** 2.419	4.359 0.352	0.000 0.725
NR does not homogeneously cause PREC PREC does not homogeneously cause NR	2.332 0.995	0.281 −1.586	0.779 0.113
URB does not homogeneously cause PREC PREC does not homogeneously cause URB	4.126 *** 2.603	2.831 0.686	0.005 0.493
GDP does not homogeneously cause PREC PREC does not homogeneously cause GDP	3.716 ** 1.748	2.250 −0.529	0.024 0.597
EC does not homogeneously cause PREC PREC does not homogeneously cause EC	3.129 1.947	1.315 −0.279	0.189 0.781
URB does not homogeneously cause NR NR does not homogeneously cause URB	3.061 2.444	1.316 0.450	0.188 0.653
GDP does not homogeneously cause NR NR does not homogeneously cause GDP	2.272 3.328	0.206 1.621	0.837 0.105
EC does not homogeneously cause NR NR does not homogeneously cause EC	2.620 2.935	0.623 1.036	0.533 0.300
GDP does not homogeneously cause URB URB does not homogeneously cause GDP	4.185 *** 5.351 ***	2.908 4.553	0.004 0.000
EC does not homogeneously cause URB URB does not homogeneously cause EC	1.338 4.843 ***	−0.684 3.661	0.531 0.000
EC does not homogeneously cause GDP GDP does not homogeneously cause EC	3.420 * 9.574 ***	1.722 10.076	0.085 0.000

Note: All variables are in logarithm forms; ***, **, and * indicate 1%, 5%, and 10% levels of significance, respectively; TEMP = Average temperature, PREC = Precipitation, NR = Natural resources, URB = urban population, GDP = GDP per capita, EC = Total energy use.

presents the results of the panel causality test, based on [82]. The test shows that bidirectional causality exists between temperature and GDP per capita; between GDP per capita and urban population; and between energy consumption and GDP per capita. Moreover, the results reveal that precipitation, NR exploitation, urbanization, and energy consumption have unidirectional causality on temperature. It is also noted that both GDP per capita and urbanization exert a unidirectional causality on precipitation. Lastly, urbanization and energy consumption have a unidirectional causal relationship, with causality running from the former to the latter.

5. Conclusions

This study explored the asymmetric effects of NR exploitation on climate change in resource-rich African countries from 1980 to 2022. In line with the extant literature, the average temperature and precipitation were used as proxies for climate change, while total NR rent was used as a proxy for NR exploitation. Other control variables used in the study are urbanization, economic growth, and total energy consumption. All the variables were adjudged and cointegrated by the [90] test for cointegration. The choice of the long-term estimation method for this study was motivated by the need to address the various econometric challenges often faced in panel data estimation. As such, the panel data of 10 resource-rich African countries were estimated using DCCE and DSUR estimation techniques. These methods are reputed for their ability to tackle the problems of heterogeneity, endogeneity, and CSD in panel data. The causal relationship among the key variables was also evaluated using the [30] panel causality test.

The study’s findings reveal a significant, asymmetric impact: both increases and decreases in NR exploitation contribute to rising temperatures and declining precipitation, with increases exerting a more pronounced effect. These results underscore the dual challenges and opportunities posed by NR management in the context of climate change. This research outcome highlights the critical role of resource-rich African countries in global climate dynamics due to their contributions to both GHG emissions and NR production. These countries face unique challenges in balancing economic development with sustainable environmental practices. Furthermore, this study demonstrates that urbanization can mitigate climate impacts through increased environmental awareness, while economic growth and energy consumption exacerbate temperature increases and precipitation declines, particularly in the context of NR exploitation.

From a policy standpoint, therefore, it can be concluded that NR exploitation, whether increasing or decreasing, has a deleterious effect on climate change in resource-rich African countries. To address this challenge, it is imperative to implement sustainable NR extraction practices, adopt green technologies, and foster collaboration among stakeholders in the NR exploitation and renewable energy value chains. Specifically, governments of resource-rich African countries need to introduce and enforce environmental sustainability legislation in the NR extraction sector to prohibit the use of illegal, unconventional, and unsustainable practices. The legislation must mandate the MNCs to always adhere to internationally acclaimed environmentally sustainable best practices and fully comply with all local regulations designed for environmental protection. Moreover, corporate bodies in the NR value chain should be urged to align their corporate social responsibility with the promotion of ecological innovation and eco-friendly practices in the sector, while tax incentives should be applied to encourage corporate bodies that are compliant.

This study has limitations that must be acknowledged. First, the study’s emphasis on Africa might limit the generalizability of its findings to other resource-rich regions with differing economic, social, and environmental contexts. Second, while the study explores the role of advanced green technologies in mitigating negative externalities from NR exploitation, it does not delve deeply into the feasibility, cost, and implementation challenges of these technologies in the African context, where resource and infrastructure constraints are prominent. Addressing these limitations in future research would enhance the comprehensiveness and applicability of the findings.