Study On The Evolution Models Of The Relationship Between Carbon Dioxide Emissions And Economic Growth

Recently,with the rapid development of global economy,carbon dioxide emissions increasingly received extensive attention of the world.The relationship between carbon dioxide emissions and economic development also becomes one of the hot issues of academic research.Many previous papers usually aim at testing the existence of the Environmental Kuznets Curve(EKC)hypothesis or study the common relationship models with traditional regression methods.As the research continues,some studies draw the seemingly contradictory conclusions,while some scholars point out that there is no universal model and both time variation and different observed objects have effect on the relationship.However,few studies further investigate the evolution models of the relationship between carbon dioxide emissions and economic development and how to select a proper model for a certain object during a specific time period.To solve the above-mentioned problems,this paper proposes a novel data-driven method,symbolic regression,to investigate the evolution models of the relationship from temporal dimension and spatial dimension.Different from traditional regression methods requiring prior assumption of model forms,symbolic regression can automatically discover the models(including model parameters and the form)from the perspective of data without any assumption of the forms,thus reducing the deviation caused by artificial selection of the models or hypothesis.Firstly,in the part from temporal dimension of this study,different from previous studies that examine the sample data of an object over a particular time period,three categories of data fragments are generated from the sample data from 67 countries during 40 years over the period 1971 to 2010,which consists of 20-years,30-years and 40-years of data,respectively.There are 14 models discovered by symbolic regression emerge after pruning the results by three thresholds,including not only three common models,monotonically increasing,inverted U-shaped and N-shaped model,but also the M-shaped model with a better applicability.Then this paper further analyzes the evolution models changing over time,and the results confirm the important impact of time variation on the relationship.Secondly,in the part from spatial dimension,unlike existing research aiming at testing whether the relationship of an object over a certain period satisfies a common model,this paper tests the applicability and evolution of the four significant models(monotonically increasing,inverted N-shaped,inverted U-shaped,and M-shaped)in countries of different areas or different economic development levels and the characteristics of the evolution model of different countries are excavated.Moreover,the comparison results with other researches verify the credibility and validity of this study.The main findings of this study are summarized as follows:(1)The impact of time variation on the relationship between carbon dioxide emissions and economic growth cannot be ignored.The relationship is confirmed to be evolutionary over time and there is no model can be widely applied in every country or during all the observed time period.(2)Model selection play an important role in the accuracy of the research results.The relationship of a specific country can be described by various models,while as the observed time period extended,both the complexity of the optimal fitting model and the applicability of the more complex model will increase accordingly.(3)Except South Asia and North America,there shows no obvious regional characteristics of the relationship.However,the relationship has high correlation with economic development level.For the countries with low level of economic development,the monotonically increasing model is the optimal one whatever the observed time period changes.(4)As for the high-income countries,the relationship between carbon dioxide emissions and economic growth illustrates a fluctuation trend.When the study period is relatively short,the inverted U-shaped model performs well,and when study period extends to 40 years or more,the applicability of inverted N-shaped and M-shaped model will be better.Especially for the high-income OECD countries,the M-shaped model has an absolute advantage over other models.Based on the above research conclusion,this paper puts forward several policy recommendations that helps countries to effectively balance the relationship between carbon dioxide emissions and economic development.Meanwhile,some management implications on China’s carbon dioxide emissions reduction are summarized from the perspective of international comparison.