An improved production-theoretical approach to decomposing carbon dioxide emissions

Production-theoretical decomposition analysis (PDA), built on production theory and data envelopment analysis, has been widely used to quantify the factors that drive CO₂ emission changes to support policy analysis and making. Existing PDA methods are usually linked to Shephard distance function and Malmquist productivity index. However, decomposition results associated with these methods may be biased and incomplete. The challenges with these methods mainly stem from the problems associated with underestimating disaggregated efficiencies and the infeasibility of linear programming. This paper proposes a modified PDA approach based on a non-radial directional distance function and global Malmquist-Luenberger productivity index. This new approach addresses the problems associated with conventional PDA methods. To show the usefulness of the proposed approach, we apply it to study CO₂ emissions in China and use the bootstrap method to test the statistical significance of the estimated results.