Efficiency of Water Provision Service: A Visual Study of Data Envelopment Analysis

Efficiency analyses of public services, such as water resource management, is in vogue owing to investments financed by taxpayers. However, in regulated sectors, i.e., without competition, data envelopment analysis (DEA) results for decision-making are questionable. Indeed, DEA is sensitive to the data, inputs, and outputs chosen by the researcher and can be highly influenced by outliers. There is no space for specialization: a decision-making unit (DMU) may be inefficient generally but efficient in one particular activity (output). Previous studies on DEA of water efficiency use second-stage analysis with exogenous factors, i.e., not controlled by management, making empirical applications unfeasible. In addition, models do not reflect changes in the specialization of DMUs through time caused by financial crises or lack of infrastructure. DEA visualization combines the standard DEA analysis with multivariate statistical methods. In this chapter, water supply and quality in 31 provinces of China are benchmarked for 2020. The results show nine efficient provinces for the traditional DEA model (all the variables). Other provinces become efficient when considering infrastructure for water supply versus water quality (few pollutants). Two provinces are efficient due to a surplus of natural water resources (overcapacity) in relation to their population, allowing for high-quality water provision, which may be transferred to regions with high water scarcity.