SVR-DEA model of carbon tax pricing for China's thermal power industry.

To mitigate the adverse effects of global climate change, the carbon tax has been gradually recognized as an important economic means to reduce carbon emissions. This paper therefore aimed to investigate the carbon tax pricing for China's thermal power industry and proposed a provincial increasing block carbon tax (IBCT) policy. By designing a forecast-optimized framework with support vector regression (SVR) and data envelopment analysis (DEA), the pricings of IBCT and flat carbon tax (FCT) were calculated. Meanwhile, the effects of both them on emission reduction were compared. The results showed that: (1) China's overall electricity demand will continue to increase in 2020, with southern and northern provinces showing stronger increases than other provinces. (2) The marginal abatement cost of each region was calculated, thus gaining an optimal three-stage form of IBCT. (3) The comparison indicated that the emission reduction efficiency of the IBCT was 23.1% higher than the FCT under the premise of equal emission reduction. The study suggests that IBCT is a more efficient type of carbon tax policy compared to FCT. Implementing IBCT can be conducive to achieving the dual goals of reducing cost burden and carbon emission in China's thermal power industry.