Abstract: Since China suffered the wide range of continuous fog haze weather frequently, the research of the cointegrating relationship between carbon emissions and economic growth has become a hot topic. As carbon emissions and economic developed rapidly in China, China is now the world's second-biggest economy and the biggest carbon emitter. It is of great significance to study the relationship between carbon emissions and economic growth. This study revisits the cointegrating relationship among carbon emission, energy consumption, economic growth and industrial structure change for China using the LM unit root test and the Gregory-Hansen cointegration test with a view of testing the environmental Kuznet's curve hypothesis in the presence of possible regime shift in long run relationship of the variables for the period 1978 to 2012 with the support of GAUSS 8.0. The problem of multicollinearity will likely result in distortion of model estimates using the least squares method directly. We used the ridge regression method to fit the model with the support of SPSS 18. The results showed that: (a) According to the results of ADF, KPSS and LM unit root test, carbon emissions, economic growth, energy consumption and industrial structure were unit root process sequences which is a prerequisite for cointegrating analysis. (b) The Gregory-Hansen cointegrating test showed that there indeed exists a cointegrating relationship among carbon emission, energy consumption, economic growth and industrial structure change in the case of structural break situations, and the break point is 1997. (c) Taking 1997 as the cut-off point, we divided the sample into two stages for analysis. Due to the presence of multicollinearity problem, we use the ridge regression analysis at each stage, which is more accurate than the OLS regression. Energy consumption, economic growth have a significant increase in carbon emissions. The coefficient of Y² is positive, which indicates that China's carbon emissions Kuznets curve has a U-shaped structure. China's economic development has not brought about a turning point reduction in carbon emissions which remains increasing with the economic development now. Comparing the results of the two periods, the elasticity coefficient of the carbon emissions for INS is not significant in the first stage and 2.02 in the second stage, which means that under other conditions unchanged, the industrial structure has a significant increase compared to the previous impact on carbon emissions. Based on this, strategies of improving energy efficiency, industrial structure regulation are proposed for carbon emissions reduction to make the "APEC blue" a "new normal".

Key words: carbon emissions, structural change, Gregory-Hansen cointegration, EKC