Carbon productivity: Reexamining the quality of economic growth in China with fixed-sum CO2 emission constraint

Scientifically and accurately measuring carbon productivity is crucial for providing constructive guidance toward achieving high-quality economic growth. In this study, we propose an improved equilibrium efficient frontier data envelopment analysis (EEFDEA) model within the framework of directional distance function (DDF). Then, we combine it with the sequential Malmquist-Luenberger productivity index (SMLPI), naming it the EEFDEA-DDF-based SMLPI approach. This novel approach accounts for not only the fixed-sum feature of CO2 emissions but also the information about input excess and output shortfalls. Moreover, it effectively captures the nature of technological progress in measuring carbon productivity growth. We apply the proposed EEFDEA-DDF-based SMLPI approach to empirically study the carbon productivity growth in China from 2012 to 2021 and analyze its drivers. Further, we explore the regional differences in carbon productivity and reveal its dynamic evolution characteristic. The empirical results indicate that the carbon productivity in China showed an upward trend during 2012–2021, with an average growth rate of 4.82 %. This growth was driven by both efficiency improvement and technological progress. Jilin and Liaoning experienced declines in carbon productivity over the sample period, while the remaining 27 provinces showed positive carbon productivity growth, and the underlying driving forces varied across different provinces. Additionally, hyper-variable density makes the largest contribution to the difference in carbon productivity. Finally, we discuss some policy implications for improving carbon productivity and provide the future research directions.