Abstract
Estimating the marginal abatement cost (MAC) of CO2 emission is critical in formulating emission reduction targets and policies. Existing studies rarely emphasized the impact of random noise on MAC estimation, and downscaling the production activity is conventionally applied as the only measure for emission abatement while the possibilities of other measures, such as increase the investment, are often neglected. This paper estimates the least MAC of CO2 for Chinese iron and steel enterprises using a stochastic semi-nonparametric method which considers both inefficiency and random noise. Multiple measures including downscaling the production activity and increasing the inputs investment, are all considered for identifying the least-cost measure for reducing emissions. In addition, the strategies corresponding to adjustment on production and response to environmental regulation of each enterprise are included in the estimation, which makes it possible for identifying the upper and lower bound of MACs. Empirical results indicate that i) the stochastic semi-nonparametric method provides a more consistent estimates with the production process, ii) the average MAC of CO2 emissions in China's iron and steel industry ranges from 2.07 to 2395 yuan/ton, and iii) increasing labor is identified as the least-cost abatement measures for most of the iron and steel enterprises listed in China's top 500 enterprise. Policy implications have been put forward to reduce the carbon abatement cost in China's iron and steel industry.