Data Envelopment Analysis: From Foundations to Modern Advancements
By Zhichao Wang, School of Economics, University of Queensland, Australia, zhichao.wang@uq.edu.au | Valentin Zelenyuk, School of Economics and Centre for Efficiency and Productivity Analysis, University of Queensland, Australia, v.zelenyuk@uq.edu.au
Abstract
Data envelopment analysis (DEA) is a mainstream method for efficiency and productivity analysis, widely applied in numerous fields, including the healthcare sector, banking, energy generation and distribution, and cross-country economic growth analysis. In this monograph, we aim to provide a compendious overview of DEA. We start with the DEA estimators in various scenarios, such as for estimating technology, cost, revenue, profit functions and related efficiency measures, and its popular variants based on different assumptions about the shape of technology. The statistical properties and extensions on DEA, such as analysis on covariates of efficiency, are also discussed and the practical tips for computations are provided.
Data Envelopment Analysis: From Foundations to Modern Advancements
Efficiency analysis methods are widely developed and applied in numerous fields, including agriculture, banking, energy, and healthcare, among others. The two mainstream approaches to efficiency analysis include data envelopment analysis (DEA) and stochastic frontier analysis (SFA). Data Envelopment Analysis: From Foundations to Modern Advancements provides a complete overview of DEA and its major variations from a practical perspective.
The authors introduce the canonical envelopment-type estimators in the production function in Section 2, their advancements in Section 3, and the estimation of cost, revenue, and profit efficiency in Section 4. The reader is then introduced to several more advanced streams of DEA literature: the productivity indexes with DEA (Section 5), the statistical properties (Section 6), including the recent development in aggregation, bias-correction, explanation of efficiency, and the “two-stage DEA” for explaining the inefficiency (Section 7). Finally, the authors provide demonstrations of some prevalent DEA estimators in R and other software using hospital data.
