The nexus between environmental degradation and economic growth continues to generate growing interest from environmental practitioners, industrialists, and researchers. Most existing studies in Africa have investigated the relationship based on the Environmental Kuznets Curve (EKC) theory under assumptions of homogeneity and spatial independence. In contrast, we investigate the EKC theory under more realistic possibilities of country heterogeneity and spatial dependence. Accordingly, we make two contributions. First, following estimation based on a sample of 48 African countries, we perform a quadratic regression for each country to account for heterogeneity. Second, we test and control for spatial dependence using the Global Moran's I test and the Maximum Likelihood Estimator (MLE) within the Fixed and Random effects on the Spatial-Durbin-Model, respectively. We also estimate the relationship using pooled OLS, Fixed and Random effects, and the generalised methods of moments (GMM). We document three key results: (1) the EKC hypothesis holds for the entire sample of 48 countries, even though the relationship is weak, (2) the relationship is sensitive to factor heterogeneity, with the EKC holding in some countries, while it breaks in others, and (3) there exist significant direct and spillover effects in the Co2-growth nexus across countries. Our findings provide a strong case for increased technological progress in pollution abatement, more abatement intensity, and adoption of cleaner production techniques. Specifically, we urge governments, multilateral organisations, and private investors to increase investments in renewable energy development projects. Given heterogeneity effects, we call for country specific measures which speaks to the Paris agreement.

环境退化和经济增长之间的联系继续引起环境从业者、工业家和研究人员越来越多的兴趣。非洲现有的大多数研究都基于环境库兹涅茨曲线 (EKC) 理论在同质性和空间独立性假设下调查了这种关系。相比之下，我们在国家异质性和空间依赖性的更现实可能性下研究 EKC 理论。因此，我们做出两个贡献。首先，根据对 48 个非洲国家样本的估计，我们对每个国家进行二次回归以说明异质性。其次，我们使用 Global Moran' 测试和控制空间依赖性 s I 分别在 Spatial-Durbin-Model 的固定和随机效应中检验和最大似然估计量 (MLE)。我们还使用池化 OLS、固定和随机效应以及广义矩方法 (GMM) 来估计关系。我们记录了三个关键结果：（1）EKC 假设适用于 48 个国家的整个样本，尽管这种关系很弱，（2）这种关系对因素异质性很敏感，在某些国家，EKC 成立，但它打破了在其他方面，以及 (3) 各国之间的二氧化碳增长关系存在显着的直接和溢出效应。我们的研究结果为增加污染减排技术进步、提高减排强度和采用清洁生产技术提供了强有力的理由。具体而言，我们敦促各国政府、多边组织、和私人投资者增加对可再生能源开发项目的投资。鉴于异质性影响，我们呼吁采取与《巴黎协定》相关的针对具体国家的措施。