The relationship between industrial structure and carbon intensity at different stages of economic development: an analysis based on a dynamic threshold panel model

Achieving the win-win goal of economic development and carbon intensity reduction, especially through industrial restructuring, is a challenge involving uncertainty and complexity. Determining which industry is green and whether it should be encouraged or limited at different stages of economic development are key issues. The relationship between industrial structure and carbon intensity was systematically analyzed in 21 industrial sectors from 1971 to 2014 in eight developed countries, with different levels of economic development, using an extended dynamic threshold model. The results indicated that there is a relationship between industrial composition and carbon intensity, and the impact trajectory of industrial structure on carbon intensity can be classified into four categories: contaminated, pollution-clean, cleaning hysteresis, and enhanced cleaning. Each type of sectoral relationship between GDP and carbon intensity would change at certain economic levels. The change points for most sectors were US$ 525 and US$ 3904 GDP per capita, which represent the points at which a country enters the mid-industrialization and high-tech industrialization stages, respectively. Therefore, the government and enterprises must upgrade their industrial structure as the national GDP increases, adjust the proportion of sectors operating according to the industrial characteristics, and improve production technology through environmental regulation.