2.1. Theoretical background

The growing availability of LT products and EGs through trade creates significant opportunities to advance sustainable economic development [10], and they are crucial in the era of climate change and economic decarbonization. Imports of LT products denote the volume of low-carbon technology products imported by the countries. These imports encompass technologies such as wind turbines, carbon capture equipment, solar panels, and biomass systems. Aggregate imports of LT products are identified using the classification developed by Howell et al. [13], which encompasses 1245 five-digit HS codes. In general, green trade represents countries’ commitment to environmental protection and sustainable development by integrating environmentally friendly goods into trade [38].

EGs denote goods linked to environmental protection, encompassing pollution control, sustainable resource management, and products that have been adapted to be more eco-friendly [39]. As there is no globally standardized definition of EGs, various databases provide differing classifications. This study adopts the Environmental Goods Index definition and uses the IMF-identified six-digit HS codes covering both environmentally protective and environmentally enhanced goods [25]. Moreover, energy-related CEI refers to the annual CO₂ emissions produced per unit of energy (kg/kWh). It measures total annual CO₂ emissions, excluding those from land-use changes, expressed in kilograms of CO₂ per kilowatt-hour of primary energy consumed.

The theoretical foundation for the analysis of the impacts of imports of LT products and EGs from China to African countries on energy-related CEI can be assessed by linking to the conceptual framework of Copeland and Taylor [15], which provides a theoretical underpinning for the channels through which trade impacts environmental outcomes. It provides evidence that trade influences environmental performance through scale, technique, and composition effects. Therefore, this framework can be a valuable lens for analyzing the relationship between LT products, EGs trade, and CO₂ emissions intensity.

More specifically, the scale effect reflects the rise in CO₂ emissions driven by the growth of economic activity caused by imports of LT products and EGs. In the absence of regulations, effective policies, and domestic absorptive capacity, increased imports of these goods will expand the economy’s scale, thereby placing additional pressure on the local environment and leading to higher emission levels [11]. In the case of trade in LT products and EGs, greater trade in green goods paradoxically raises emissions [17]. Therefore, if trade growth stimulates energy-intensive production, it increases the sector’s emissions, thereby raising overall carbon emissions [11].

Conversely, the technique effect refers to the impact of technological improvements on CO₂ emissions through the importation of advanced products. When cleaner technologies are introduced, they replace older, more polluting systems, ultimately lowering carbon emissions [11], as trade can enhance availability and reduce the cost of environmentally friendly goods and technologies. This is particularly critical for economies with limited access to these goods, services, and technologies, or where domestic production capacity remains insufficient to supply them at sufficient scale or at internationally competitive prices [16]. Moreover, international trade facilitates the cross-border diffusion of environmental technologies, thereby enhancing energy efficiency and promoting environmentally sustainable production processes [40]. Also, trade-driven income growth may increase public demand for improved environmental quality, thereby mitigating emissions [16]. By facilitating greater access to goods and technologies that promote energy efficiency, trade can therefore contribute significantly to addressing environmental degradation [16].

Finally, the composition effect suggests that the environmental impact of trading these products hinges on whether a country’s specialization is concentrated in energy-intensive industries or in cleaner, low-emission sectors [17]. That is, imports of LT products and EGs can alter the allocation of economic resources across sectors with varying emission intensities, depending on a given economy’s comparative advantage [41].

2.2. Summary of empirical literature and hypotheses development

Building on the theoretical perspectives outlined above, numerous empirical studies have examined the effects of low-carbon and green technologies on environmental quality across economies, as well as the impact of trade in LT products and EGs on emissions. For instance, using the MMQR approach, Sharif et al. [42] examined the role of green technology in environmental quality and found that it positively influences environmental performance.

Also, employing a dynamic autoregressive distributed lag (ARDL) model, Ali et al. [39] investigated the impact of both green and non-green technological innovations on carbon emissions in Canada and indicated that green innovation significantly reduces carbon emissions. Using the MMQR approach, Ramzan et al. [43] investigated the impact of eco-friendly technologies on CO₂ emissions across the world’s 10 most sustainable economies and found that LT significantly reduces carbon emissions.

Similarly, Fang [44] examined the effects of green technologies on CO₂ emissions and found that LT significantly reduces pollution in China. Also, Lisha et al. [45] and Habiba et al. [46] found consistent findings for the Brazil, Russia, India, China, and South Africa (BRICS) countries and the top carbon-emitting countries, respectively. However, Sakilu and Chen [47] found that green technological innovations have a positive effect on CO₂ emissions in the 19 countries with the highest emissions.

Furthermore, some empirical work has evaluated the effects of EGs trade on CO2 emissions, though energy-sector-specific investigations are limited. For example, Can et al. [20] investigated the role of green goods in the environmental performance of Organization for Economic Co-operation and Development (OECD) economies and confirmed that environmentally friendly goods contribute to environmental sustainability. Also, Alvi et al. [19] showed that imports of EGs in developing economies decrease emissions. Likewise, Herzer [25] implied that foreign green technology lowers CO₂ emissions in Group-7 (G7) economies. The findings also showed that spillovers of foreign environmentally friendly technology into domestic CO₂ emissions occur through the import of EGs.

Moreover, Dong et al. [11] argued that increased green product exports can effectively mitigate CO₂ emissions in China. Similarly, Wei et al. [48] examined the effects of green trade and other determinants on environmental quality in the world’s top green future, and the findings showed that green trade significantly improves environmental quality.

Unlike the aforementioned empirical works, some studies argue that trade in EGs is unable to mitigate CO₂ emissions. For instance, Bai et al. [23] analyzed the effect of green trade on CO₂ emissions within G7 nations. Their results revealed that expanding the share of environmentally friendly trade may be counterproductive, as it contributes to higher consumption-based carbon emissions. Similarly, Hu et al. [24] suggested that trade in EGs does not necessarily yield environmental benefits in the absence of complementary policies, as the diverse end uses of such goods can increase energy consumption and, consequently, environmental degradation. Also, Liu et al. [10] and Zugravu-Soilita [17] showed that trade in EGs alone is ineffective at addressing environmental problems. Likewise, Liu et al. [49] indicated that EGs exports are detrimental to China’s green development. Based on the theoretical and empirical works discussed, the following hypothesis was developed.

Hypothesis 1 (H1): Imports of LT products and EGs from China negatively affect energy-related CO₂ emissions in African countries.

Moreover, it is argued that the effectiveness of LT products and EGs in mitigating carbon emissions may be constrained by recipient countries’ economic capacity and the regulatory performance of importing countries [30]. Therefore, the effects of LT products and EGs trade on CO₂ emissions can be influenced by differences in economic development and IQ in trade partner countries. In line with this justification, Ha [50] studied the impact of EGs trade on environmental quality, distinguishing between countries with different income levels, and showed that EGs trade improves environmental quality, particularly in countries with strong and advanced institutional systems. Also, using data from 46 countries, Li et al. [51] assessed the effects of digital trade on carbon emissions, controlling for countries’ heterogeneity, and found that digital trade has a significant mitigating effect on emissions, though its impact varies across countries at different levels of development.

Moreover, IQ shapes trade composition and facilitates a shift toward cleaner and more environmentally sustainable products [52]. Thus, stable political environments and higher IQ are conducive to the adoption of LT products and EGs, as well as to the expansion of trade. Chen et al. [53] argued that IQ increases the adverse effects of green innovation on energy intensity. Thus, IQ is assumed to mitigate energy-related CEI by strengthening countries’ domestic absorptive capacity to optimize the emissions-reduction benefits of imports of LT products and EGs. Therefore, based on this foundation, the following hypothesis is articulated.

Hypothesis 2: The role of LT products and EGs from China on energy-related CEI will be influenced by income and IQ heterogeneity of African countries.

2.3. Literature summary and research gaps

First, existing empirical work has investigated the effect of EGs on aggregate CO₂ emissions and other environmental performance indicators [10,11,16,19,20,23,25,48]. However, these studies largely overlook the connection between LT products and the EGs trade and energy-related carbon emissions. Hence, investigating how China’s LT products and EGs trade affect the sectoral level, with more carbon emissions contributing to the energy sector, offers valuable insights that enrich the current body of literature on sustainable resource utilization and green finance.

Second, although empirical studies have examined the effects of EGs trade on environmental sustainability [19,23,48], they focus on developed economies or samples of emerging countries, omitting African countries and the influence of trade in LT products and EGs on energy-sector carbon emissions.

Third, prior empirical findings on the link between EGs and CO₂ emissions are inconclusive. Some studies have reported the CO₂-emissions-mitigation roles of EGs [11,19,25], while others have documented a positive or insignificant relationship between EGs and CO₂ emissions [10,17,23]. Moreover, empirical work on the relationship between LT products and CO₂ emissions from the energy sector is scarce.

Fourth, from an econometric standpoint, the findings of this study have significant methodological implications, particularly given endogeneity concerns and considerable heterogeneity. This necessitates adopting more advanced econometric approaches capable of addressing such complexities. To this end, the study employs IV-GMM and MMQR with fixed effects, which effectively address endogeneity and heterogeneity concerns.

Lastly, acknowledging the diversity among African countries in terms of IQ and income levels, this research extends beyond prior empirical works by analyzing how China’s LT products and EGs trade affect energy-related CEI in Africa while controlling for income and IQ heterogeneity. Besides, it assesses the role of regional integration in the association between imports of LT products and EGs from China and energy-related CEI in Africa. Consequently, the present study contributes a novel perspective that enriches existing research by providing more concrete evidence on how China’s LT products and EGs trade affect CEI in African countries.

3.1. Theoretical base for model construction and data

The assessment of the relationship between the target variables (China’s LT products and EGs exports to African countries) and energy-related CO₂ emissions intensity can be linked to Ehrlich and Holdren’s [54] and Commoner’s [55] theoretical background and the subsequent economic modeling, which has been extensively applied in the STIRPAT analysis [56]. Besides, the econometric model used in this study can be specified within the Environmental Kuznets Curve (EKC) framework, as both GDP per capita and its square are included.

Therefore, using energy-related CEI as the dependent variable, we follow a STIRPAT model and the EKC hypothesis, including per capita GDP and its square term to capture economic growth, population (PO), low-carbon technology product imports from China (LTPI), environmental goods imports from China (EGI), and other regressors. Therefore, the energy-related CEI defined in Eq (1) below is expected to be negatively affected by LTPI (). Likewise, CEI in Eq (2) will be negatively influenced by EGI ().

(1)

Based on theoretical grounds, the quality of institutions (QI) and the square of GDPpc are expected to negatively contribute to energy-related CEI (). However, GDP per capita of countries (GDPpc), industrialization (IND), urbanization, and population size will positively influence energy-related CEI (). Finally, the effects of FDI on energy-related carbon emissions can be positive, subject to the pollution haven hypothesis ().

(2)

3.2. Data

This research uses unbalanced panel data on imports of LT products and EGs from China by 42 African countries for the period 2003–2023, based on data availability for the major variables. (Sample countries: Algeria, Angola, Benin, Burkina Faso, Botswana, Burundi, Cameroon, Cabo Verde, Central African Rep., Comoros, Congo Rep., Côte d’Ivoire, Egypt, Ethiopia, Eswatini, Ghana, Gabon, Gambia, Guinea, Kenya, Lesotho, Madagascar, Malawi, Mali, Mauritius, Mauritania, Morocco, Mozambique, Namibia, Niger, Nigeria, Rwanda, Senegal, São Tomé and Príncipe, Seychelles, South Africa, Tanzania, Tunisia, Togo, Uganda, Zambia and Zimbabwe). Energy-related CEI is the dependent variable, represented by annual CO₂ emissions per unit of energy (kg/kWh), excluding emissions from land-use change. Data on energy-related CEI are sourced from the WB database.

Besides, there are two target variables: LT products and EGs. LT products are the amount of low-carbon technology products imported from China by each sample African country, based on data from the IMF Climate Dashboard. These imports include wind turbines, solar panels, carbon capture equipment, and biomass systems [57]. The IMF adopts a top-down framework that applies a standardized methodology and uniform definition across a broad set of economies. This classification enables comprehensive assessments of aggregate bilateral trade trends in LT products. Nevertheless, because the framework does not categorize individual LT products according to specific industries or sectors, it remains limited in facilitating disaggregated sectoral analysis [58].

China’s LT products exports to African countries are the aggregate of all LT product imports from China reported by each African sample economy that meet the definition of LT products. Aggregate LT products exports are constructed by using the classification presented in Howell et al. [13]. These comprise a list of harmonized system (HS) codes (HS 2017) and are a subset of EGs [59], which includes 124 codes constituting LT products in industry supplies, capital goods, transport equipment and consumer goods [59], consisting of energy-efficient appliances, renewable energy goods, electric vehicles, energy storage solutions and smart grid systems [60].

EGs represent imports from China to African countries, drawn from the IMF database, covering a broad range of technologies and materials that support sustainability and cleaner production practices. We use the EGs definition and data compiled by the IMF, which include goods directly associated with environmental protection, such as products used for pollution control and resource management, as well as those modified to be more environmentally friendly [59]. These goods represent an aggregation of HS 6-digit commodities classified as EGs by the International Monetary Fund (IMF) [25], including environmental protection goods across various segments, such as food and beverages, industrial and primary industrial supplies, fuels, transport equipment, capital and consumer goods [59]. The patterns in the link between the target variables and energy-related CEI are presented in Fig 3.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. The patterns of LT products, EGs and energy-related CEI.

https://doi.org/10.1371/journal.pone.0339433.g003

Moreover, based on the theoretical and empirical works, several control variables are included in the analysis. GDP per capita, GDP per capita squared, FDI, IQ, population size, and urbanization are included to control for economic, institutional and demographic factors that are identified as the main factors affecting CO₂ emissions, subject to the STIRPAT framework, designed to include economic, demographic and other factors in the CO₂ emissions analysis [18]. GDP per capita and its squared term are included to capture the EKC hypothesis, which posits that carbon emissions rise in the early stages of growth but decline as countries reach higher income levels. Population size and urbanization are used to capture the effects of demographic dynamics on energy-sector carbon emissions.

Aggregate IQ is derived from the World Bank’s (WB’s) six governance quality indicators grouped into three dimensions: economic governance, captured by government effectiveness and regulatory quality; political governance, represented by political stability and voice and accountability; and institutional governance, measured through control of corruption and the rule of law [61]. The IQ index is included in the analysis, as it is believed to affect trade composition and foster a transition towards cleaner products [52]. Principal component analysis (PCA) is applied to derive the IQ index (Table A1). The results show that institutional governance (rule of law and control of corruption) contributes more to the aggregate IQ index, followed by economic governance (government effectiveness and regulatory quality), as they have relatively larger positive loadings (Table A1). Details on the definition, data sources, and links are discussed in Table 1.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Details about major variables.

https://doi.org/10.1371/journal.pone.0339433.t001

3.3. Method of analysis

An empirical model based on the EKC hypothesis and the STIRPAT framework is developed to test the role of LT products and EGs imported from China in the energy-related CEI of African countries, as shown in Eqs (1) and (2). However, there may be potential reverse causality between the target variables, the control, and the dependent variable. For example, increases in imports of LT products and EGs can influence energy-related CEI, while improvements in energy-related CEI can also affect countries’ trade performance. This bidirectional relationship may give rise to simultaneity bias, potentially yielding inconsistent and biased estimates when estimated using ordinary least squares (OLS) or standard panel regression techniques [60,62]. To address reverse causality and endogeneity concerns, the IV-GMM approach, a robust approach for controlling for potential reverse causality, is applied [63].

This approach mitigates concerns arising from endogeneity, omitted-variable bias, and measurement error, thereby ensuring more reliable and consistent results [62,63]. It also accommodates potential autocorrelation and heteroskedasticity issues. This estimator is particularly appropriate for panel datasets with a larger cross-sectional dimension than the time dimension (N = 42 > T = 21). Moreover, even in the presence of uncertain heteroskedasticity, the IV-GMM effectively mitigates omitted-variable bias, producing consistent and efficient estimates [60]. Therefore, the IV-GMM approach proposed by Baum et al. [63] is applied to estimate Eqs (3) and (4). We employ lagged values of low-carbon energy as internal instruments, given the difficulty of identifying valid external exogenous instruments [60,64]. To test the validity of the instruments, the Kleibergen-Paap rk LM statistic and the Cragg-Donald F-statistic are used to check for underidentification and weak identification, respectively, while the Hansen J test is used to check for overidentification.

(3)

(4)

Moreover, we used the MMQR approach for the sensitivity analysis. The MMQR explicitly captures heterogeneous effects of LT products and EGs on energy-related CEI by allowing covariates to exert varying influences across the entire conditional distribution. Accordingly, this method is well suited to analyzing data that do not follow a normal distribution (see Table A7 in S1 Appendix), as it generates robust and reliable estimates even in the presence of distributional irregularities [65]. MMQR is also resilient against outliers in panel data [42,66]. It addresses the possibility of endogenous features in the response variable [43]. Thus, we employ the MMQR to evaluate the effects of LTPI and EGI on energy-related CEI using Eqs (5) and (6).

(5)

(6)

Where denotes τth conditional quantile function. X set of independent variables (LTPI, EGI, FDI, GDPpc, GDPpc², IND, UR, PO, and QI).

Furthermore, Lewbel’s [67] IV approach is applied to address reverse causality concerns, as it is well-suited to handle endogeneity arising from omitted-variable and simultaneity bias [28]. Therefore, external instruments are constructed for imports of LT products and EGs for each African country by using the average LT product imports and EGs share of other African countries in the sample, as specified in Eqs (7) and (8) following the work of Owusu et al. [28]. Fig 4 presents a comprehensive overview of the methods used.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Methodological framework.

https://doi.org/10.1371/journal.pone.0339433.g004

(7)

(8)

Where, aveLTPI_t and aveEGI_t are the average LTPI and EGI, respectively, while LTPI_t and EGI_t are the sum of LTPI and EGI from China across the sample African countries in period t, respectively, while LTPI_it and EGI_it are country-time-specific LTPI and EGI from China, and n is the total number of countries in the sample.

4.1. Preliminary analysis

Before proceeding to the regression analysis, several preliminary statistical analyses are conducted. Table 2 reports the statistical summary of the major variables included in the investigations of the effects of LTPI and EGI on energy-related CEI, followed by correlation and multicollinearity tests. The correlation coefficients between LTPI and EGI are higher (Table A2), and multicollinearity concerns arise given their individual variance inflation factor (VIF) values (Table 3A). As a result, they are separately included in the analysis. Moreover, the slope heterogeneity (S-H) test results shown in Table A4 indicate that the slope coefficients exhibit heterogeneity.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Summary of major variables.

https://doi.org/10.1371/journal.pone.0339433.t002

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. The effects of LC products and EGs on CEI.

https://doi.org/10.1371/journal.pone.0339433.t003

4.2. Baseline results

The effects of LT products and EGs imported from China on the energy-related CEI of African countries are presented in Table 3 below. LT products and EGs are controlled separately due to the very high collinearity between them. The first two Columns (I and II) report the OLS results, and Columns (III) and (IV) present the findings of the IV-GMM approach. The diagnostic tests confirmed that the IV-GMM model is neither weakly identified, under-identified, nor over-identified. Therefore, the baseline results are interpreted using the IV-GMM findings reported in Columns III and IV.

The findings imply that GDP per capita positively affects energy-related CEI, suggesting that larger economies consume more energy and emit more carbon from the energy sector. This can be attributed to the fact that the economic growth model of most African countries has yet to transition to a low-carbon paradigm, with economic expansion continuing to generate higher carbon emissions. Moreover, the coefficient on the square term of GDP per capita is negative in all specifications. Therefore, the assumptions of the EKC framework are supported, as the coefficients for per capita GDP and its square have the expected signs. This indicates that energy-related CEI initially increases with economic expansion to a certain point, beyond which further per capita GDP growth mitigates energy-related CEI in African countries.

Moreover, FDI inflow has a significant adverse effect on energy-related CEI. Therefore, the results confirm the pollution halo hypothesis, which predicts that FDI helps reduce CO₂ emissions by facilitating the transfer of improved management practices and environmentally friendly technologies to host countries. Likewise, the aggregate IQ indicator has a negative impact on energy-related CEI, suggesting that higher IQ levels contribute to mitigating carbon emissions from the energy sector in African countries. However, population size and urbanization positively contribute to energy-related CEI in African countries. Population size and urbanization are significant factors contributing to the increase in energy-related CEI in African countries. Urbanization increases energy emissions intensity, as urban societies typically use more electricity, appliances, vehicles, and services, which may lead to greater reliance on fossil fuels in African countries. Similarly, an increase in population size increases energy demand and results in higher emissions from energy consumption.

Regarding the target variables, the findings suggest that LT products imported from China have a significant negative effect on the energy-related CEI in African countries. Specifically, a 1% increase in China’s LT product imports to African countries reduces energy-related CEI by 0.077% (Column III, Table 3). Therefore, the results indicate that African countries’ imports from China, especially environmentally related technology products, have a significant negative impact on their energy-related CEI. These results align with both theoretical and empirical arguments. The findings can be loosely argued to show that importing LT products, such as solar panels and electric vehicles, helps reduce carbon emissions by providing access to advanced, cleaner technologies that improve energy efficiency and reduce pollution [68]. This allows countries, especially African nations, to pursue economic growth without increasing energy-sector emissions. Additionally, by promoting green technologies and low-carbon goods, imports can play a crucial role in advancing sustainable development and supporting the transition to a low-carbon economy by replacing high-emission conventional energy sources. Consequently, China’s LT product imports can help these countries meet their emissions-reduction targets under the Paris Agreement.

Similarly, EGs imports from China to African countries negatively impact the energy-related CEI of African nations. More specifically, a 1% increase in EGs imported from China reduces the energy-related CEI of African countries by 0.067% (Column IV, Table 3). This can be roughly explained by the fact that trade in EGs provides economies with more opportunities to adapt environmental technologies to local needs. These goods can help transition from fossil fuel-based energy systems to cleaner, renewable sources by enhancing energy efficiency and supporting low-carbon production in African countries’ energy sectors.

4.3. Robustness Analysis

First, we run the exercises excluding countries with significant carbon emissions in the sample countries, and then exclude countries with a larger proportion of LT products and EGs imported from China. Notably, South Africa, Egypt, Algeria, and Nigeria rank among the four highest emitters in Africa. In particular, South Africa is ranked 15th among the world’s largest emitters [69]. Moreover, South Africa, Nigeria, Ethiopia, and Egypt import a larger proportion of LT products from China (see Fig A1). Therefore, the analysis is repeated, excluding these countries from the sample to further control for outliers, and the results reported in Table A5 reinforce the baseline results reported in Table 3.

Second, we further examine the effects of LTPI and EGI on total CO₂ emissions intensity (Table A6). The findings suggest that both LT products and EGs imported from China negatively affect total CO₂ emissions intensity in African countries, indicating a role for these imports in mitigating CO₂ emissions.

Third, the analysis is repeated using the MMQR approach to account for potential heterogeneity across the conditional distribution. Before running the MMQR strategy, a normality test is conducted to assess the distribution of the variables. The results indicate that the variables do not follow a normal distribution (Table A7). Consequently, a non-parametric approach, namely, the MMQR, is deemed appropriate for the analysis.

Regarding heterogeneity in energy-related CEI, LT products have a greater mitigating impact in African countries with higher carbon-emission intensity per unit of energy consumption. The absolute value of LT coefficients increases from 0.032 to 0.084, indicating that a 1% increase in LT product imports from China leads to a decrease in energy-related CEI by 0.032% to 0.084% across the 25^th to 90^th percentiles (Table 4). Additionally, the results clearly show that the reduction in energy-related CEI due to China’s EGs exports to African countries is substantial across the 25^th to 90^th percentiles, decreasing from 0.035% to 0.039% as the percentile increases. Therefore, the findings suggest that LT products and EGs imported from China provide greater benefits for reducing energy-related CEI in African countries with the highest proportions of energy-related CEI.

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. The effects of LT products and EGs on energy-related CEI (MMQR).

https://doi.org/10.1371/journal.pone.0339433.t004

Fourth, the analysis is repeated using Lewbel’s IV approach using external instruments for the target variables. The analysis outcomes are reported in Table 5 and support the results in Table 3.

Download:

• PNG
  larger image
• TIFF
  original image

Table 5. The effects of imports of LT products and EGs on energy-related CEI (IV).

https://doi.org/10.1371/journal.pone.0339433.t005

4.4. Further heterogeneity analysis

The baseline results in Table 3 indicate that imports of LT products and EGs from China negatively affect energy-related CEI in African countries. However, because the sample countries are quite diverse, these results may not apply broadly to all such economies. To address this variation, we further examine how the effects of imports of LT products and EGs from China vary across the sample countries, accounting for differences in income levels and institutional quality. First, African countries considered in the analysis are grouped into low-income and middle-income economies, and the results are presented in Table 6.

Download:

• PNG
  larger image
• TIFF
  original image

Table 6. The effects of imports of LT products and EGs on energy-related CEI (income variation).

https://doi.org/10.1371/journal.pone.0339433.t006

The findings are mixed: imports of LT products and EGs from China have a strong negative impact on energy-related CEI in middle-income economies, but a positive impact in low-income African countries. A 1% increase in imports of LP products and EGs from China into middle-income countries results in decreases of 0.155% and 0.1% in energy-related CEI, respectively. Moreover, an additional analysis is conducted for middle-income countries excluding South Africa to assess whether the country’s distinctive energy structure and status as one of the world’s largest carbon emitters may disproportionately influence the aggregate results for the middle-income country sample. The results in Table A8 imply there is no considerable skewing role. Conversely, a 1% increase in imports of LT products and EGs by low-income African countries raises energy-related CEI by 0.255% and 0.285%, respectively, suggesting that greater reliance on these imports may hinder the green energy transition in these economies.

Second, the analysis is repeated, controlling for IQ variation within the sample of African countries by dividing them into groups with lower and higher IQs. The findings in Table 7 suggest that African countries with higher IQ scores experience positive effects of LT products and EGs imports from China on energy-related CEI mitigation. However, for countries with lower IQ, the effect remains positive and statistically significant, indicating that imports of LT products and EGs from China tend to exacerbate energy-related carbon emissions rather than promote a cleaner energy transition in these economies.

Download:

• PNG
  larger image
• TIFF
  original image

Table 7. The effects of imports of LT products and EGs on energy-related CEI (IQ variation).

https://doi.org/10.1371/journal.pone.0339433.t007

6.1. Conclusions

This study examines how China’s trade in LT products and EGs influences energy-related CEI across 42 African nations from 2003 to 2023. It employs advanced econometric techniques, including IV-GMM, MMQR, and Lewbel-IV approaches. Additionally, the analysis accounts for variations in IQ and income levels among African countries. The study findings can be summarized as follows: (1) The impact of LT products and EGs imported from China on energy-related CEI of African countries is negative and helps reduce energy-related CEI across the whole sample. (2) Regarding the distributional heterogeneity of African countries’ energy-related CEI, those with the highest energy-related CEI benefit more from LT products and EGs imported from China compared to countries with lower energy-related CEI. (3) The effects of LT products and EGs imported from China on energy-related CEI vary when the income heterogeneity of African countries is controlled. LT products and EGs imported from China help mitigate energy-related CEI in middle-income African countries and positively contribute to energy-related CEI in low-income economies. (4) The findings suggest that LT products and EGs imported from China help reduce energy-related CEI only in African countries with better QI and exacerbate energy-related CEI in African countries with weak QI.

6.2. Policy implications

Based on the study’s findings, the following policy implications are proposed:

First, our findings suggest that imported LT products and EGs from China help reduce energy-related CEI by supporting the inflow of green products. Therefore, policymakers should work to ease trade restrictions on LT products and EGs to encourage the diffusion of these innovations across African markets. This can be achieved through reducing tariff and non-tariff barriers to trade in LT products and EGs by applying policy measures such as lowering tariffs on green products, improving customs and border-clearance procedures, and harmonizing technical and certification standards.

Second, governments in trade partner African countries and China should develop a robust policy framework to facilitate knowledge sharing, joint ventures, and capacity-building initiatives that help recipient countries effectively use and maintain imported low-carbon technologies through LT products and EGs

Third, the heterogeneity analysis reveals that African countries with higher energy-related CEI quantiles derive greater mitigation benefits from importing LT products and EGs from China, underscoring the need for differentiated carbon mitigation strategies tailored to the carbon-intensity profiles of individual African economies. For example, highly carbon-intensive African economies such as South Africa and Nigeria should prioritize importing Chinese LT products and EGs to decarbonize the energy sector more than lower-emission economies. Therefore, these differentiated approaches would enable African economies to align green technology imports with their specific carbon-intensity profiles and development needs, thereby enhancing the effectiveness of energy-related carbon mitigation policies.

Fourth, the study’s findings underscore the critical importance of strong QI in facilitating energy-sector decarbonization through the imports of LT products and EGs from China. More specifically, in African countries with stronger QI, where imports of LT products and EGs effectively reduce energy-related CEI, governments should reinforce transparent regulatory frameworks and governance, and deepen trade with China to accelerate clean energy adoption and the diffusion of technology. In African countries with weak QI, where EGs and LT product imports may contribute to rising energy-related CEI, governments should prioritize institutional reforms and strengthen regulatory enforcement to translate these imports into meaningful carbon mitigation outcomes.

Fifth, the mixed effects of imports of LT products and EGs from China on energy-related CEI suggest the need for differentiated policy responses based on income levels to ensure gains from green trade and to effectively mitigate energy-related CEI. In particular, in middle-income economies, where LT products and EGs imports contribute to reductions in energy-sector CEI, governments should strengthen green energy policies and enhance trade partnerships with China to accelerate the diffusion of technologies and energy-efficiency gains from these imports. Conversely, in low-income countries, where such imports may inadvertently increase CEI, policymakers should prioritize institutional development, technical capacity building and supportive energy infrastructure to benefit from LT products and EGs imports.

Finally, a stronger commitment to the African Continental Free Trade Area (AfCFTA) is essential, as deeper regional integration can boost intra-African trade, reduce tariffs and non-tariff barriers, improve logistics and energy infrastructure, and lower transaction costs, thereby accelerating structural transformation. This is particularly crucial for low-income African countries, enabling them to better leverage imports of LT products and EGs from China to mitigate energy-related CEI. Moreover, governments should establish regional green value chains, alongside improved customs efficiency and regulatory coordination aligned with the AfCFTA, to further enhance the emissions-reducing potential of green imports and support a continent-wide low-carbon transition.

6.3. Limitations of the study

The scope of this study is limited to examining the impact of LP products and EGs imported from China to African countries on energy-related CO₂ emissions. Future research should undertake comparative analyses of the effects of LT products and EGs trade across different regions. Although the IV-GMM and MMQR approaches employed in this study effectively address endogeneity and distributional heterogeneity, they do not explicitly capture the short- and long-run effects of imports of LT products and EGs on energy-related CEI; these effects can be evaluated separately in future studies. Also, the investigation could be expanded to assess the role of LT products and EGs trade on energy-related CEI in both developing and developed countries, given the differences in technological, economic, and emissions patterns. Finally, for low-income economies, the relationship between green goods trade and energy-related emissions should be further empirically examined to assess whether complementary conditions or rebound effects influence the overall impact. The author used ChatGPT to correct typos and improve language during manuscript preparation. Afterward, the authors carefully reviewed, edited, and approved the final version. The authors assume full responsibility for the accuracy, content, and integrity of the published work.