Unraveling the pollution Haven effects of goods and services trade in African countries

In the context of 47 African States, we investigated the distinguishing effect of the goods and services trade in generating pollution from 2000 to 2022. The study relied on two prominent panel data estimation techniques, random effect and system GMM. Findings show that positive scale and negative technique effects exist in the African continent, and this affirms the environmental Kuznets hypothesis. A rise in foreign trade as disaggregated into goods and services trade, increases CO2 emissions, while we failed to reject the hypothesis that “goods trade is more polluting in terms of generating CO2 emissions than services trade”. We observed a nonlinear link between goods trade and CO2 emission in the general sample of African countries and the sub-sample of resource-rich African countries, while in the case of services trade and CO2 emission nexus, a non-linear link is only observed in the sub-sample of resource-poor African countries. The observed threshold of goods trade in the general sample is 6.06, which is yet to be attained, and the 4.46 threshold in the resource-rich countries, which has been attained. The threshold of services trade is 5.56 in the resource-poor countries. This threshold has been attained by resource-poor countries. Other pollutants (GHG, PM2.5, N2O, and F-gases) all show a non-linear link with services trade with thresholds of 10.25, 3.13, 159453368.8, and 859.85. Goods trade-GHG and goods trade-F-gases show an inverted U-shaped threshold of 2,568,702,300 and 136.87, and a U-shaped threshold of 29.36 with N2O emissions. The premise that trade openness changes the industrial composition toward polluting sectors was validated in most cases. From our estimate using CO2 emissions, energy held the second dominant role in exacerbating pollution. In the general sample and sub-sample, we reject the hypothesis that trade may bring the right energy-efficient technology that lowers pollution. When other pollutants are used, we have established, to some extent, evidence of this hypothesis. In the case of CO2, our findings also validate the pollution haven and the factor abundance effect of both goods and services trade. When other pollutants are used, a pollution haven in services trade exists in the case of PM2.5 and F-gases, and PM2.5 and N2O for goods trade. In the case of CO2 emissions, although trade is polluting, the use of trade to transfer pollution from the developed world to African countries is even more polluting. The study recommends the need to implement environmental regulations and policies to curb pollution transfer to the African continent.