Cluster analysis of the relationship between carbon dioxide emissions and economic growth. (10th July 2019)

• Record Type: Journal Article
• Authors: Li, Wenli; Yang, Guangfei; Li, Xianneng; Sun, Tao; Wang, Jianliang
• Is Part Of: Journal of cleaner production. Volume 225(2019)
• Page Start: 459
• Page End: 471
• Publication Date: 2019-07-10
• Subjects: Cluster analysis -- Carbon dioxide emission -- Economic growth -- Symbolic regression -- Apriori; Factory and trade waste -- Management -- Periodicals; Manufactures -- Environmental aspects -- Periodicals; Déchets industriels -- Gestion -- Périodiques; Usines -- Aspect de l'environnement -- Périodiques; 628.5
• DOI: 10.1016/j.jclepro.2019.03.220
• Languages: English
• ISSNs: 0959-6526

Abstract

Abstract: As global warming continues to worsen, the balance between carbon dioxide emissions and economic growth has received increasing attention and carbon-reduction comes to be an urgent task in many countries. In literature, various regression models have been developed to investigate the relationship between carbon dioxide emissions and economic growth, such as the inverted U-shaped EKC model, inverted N-shaped model, etc., which play critical roles in analyzing the relationships. Existing studies suggest that some countries follow similar models to describe the relationships, while others employ different ones. Regarding the interplay between carbon dioxide emissions and economic growth, there lacks a cluster analysis to systematically uncover the similarity of countries that employ models more similar to each other than to the countries in other clusters. In this paper, a novel clustering approach is proposed to identify clusters of 67 countries from the spatial, temporal, and descriptive dimensions. Unlike the traditional clustering technique, in which clusters are determined by geometric distances, the clusters in this research are obtained based on the differences between the fitting models of the countries in each cluster. The first step of the approach is to find clusters of countries sharing similar models in a given year based on the symbolic regression method, and the second step is to determine the countries that frequently cooccur in the same cluster by the Apriori algorithm. The results present two high-order clusters with different dynamic features during the period between 1971 and 2010. One high-order cluster mainly consists of countries in higher level of income while the other contains lower income level countries, and the carbon intensities of the two high-order clusters have distinct differences. The findings suggest that the countries within the same cluster could learn more lessons from each other, while countries belonging to different clusters should not be examined together indiscriminately. Several policy implications are provided, which may inform decision-making for policymakers when choosing proper learning objects and help researchers with designing optimal models for specific countries. Highlights: ● Conduct a cluster analysis of the nexus between CO2 emissions and economic growth. ● A novel data-driven clustering approach based on symbolic regression is proposed. ● Intelligently discover the underlying models and extract the dynamic clusters. ● Countries that frequently cooccur in the same cluster are identified. ● Environmental policymaking is not "one size fits all" and clusters will be helpful.