RESUMO
To promote sustainable water resource management, the Chinese government has enacted restrictive policies and targets in terms of total water usage, water intensity, and pollution emissions. While data envelopment analysis (DEA) has been extensively adopted in assessing industrial water resource systems, previous studies have not integrated the realistic constraints on total water consumption and total pollution emissions into a unified framework. This paper examines the system as a two-stage process of water use (WU) and water pollution abatement (WPA), where fixed-sum constraints are imposed on both water consumption and pollution emissions. To address such two-stage DMUs with fixed-sum inputs and fixed-sum outputs, we propose a novel two-stage fixed-sum DEA approach and apply it to evaluate the performance of industrial WU-WPA systems for 29 provinces in China from 2014 to 2018. The results are as follows. From the overall efficiency perspective, the industrial WU-WPA system in China is overall efficient, but provincial overall efficiency polarizes with 19 provinces evaluated as overall inefficient. From a time perspective, overall efficiency shows an increasing then decreasing or descending trend nationally and across the 16 provinces; on the national average, WPA efficiency shows an increasing and then decreasing trend, while WU efficiency demonstrates consistent improvement over time and surpasses WPA efficiency after 2016. In terms of the four areas, regional disparities in the overall efficiencies are converging; for both the whole system and the sub-stage, the eastern area performs the best, followed by the western, northeastern, and central areas. Based on the empirical results, suggestions for improving industrial water resource management are given at the national, regional, and provincial perspectives respectively.
