Estimating output-specific efficiencies (Q1396173)

The book addresses the issues of output-specific efficiency measurements. Prior to this work, no technique capable of handling this type of problems has been available. The book successfully fills this gap by providing a statistical foundation for output-specific efficiency estimation; however the author himself admits that further improvements and generalizations are required before the method is established as a standard tool. The proposed technique is essentially a two-step method that starts with a frontier estimation procedure which is followed by an output-specific efficiency estimation procedure. The book consists of ten chapters organized in four parts. Part 1 (Motivating the concept) can be seen as an introduction to the whole monograph. It gives a description of the problem area illustrated by examples, a brief review of the main techniques widely used in the book: Data Envelopment Analysis (DEA), Stochastic Frontier Analysis (SFA) and Markov Chain Monte Carlo (MC2) techniques. Part 2 (Operationalizing the concept) consists of Chapters 2--7 and takes up most of the book's volume. Chapter 2 presents various adaptations of DEA to handle frontier estimation in the presence of output-specific efficiencies. Chapter 3 discusses the conjecture that the marginal output-specific efficiency distributions are limits of the radial efficiency distributions. The \(\text{MC}^2\) technique is described in Chapter 4. The statistical background for output-specific estimations is provided in Chapter 5. The identification issues based on graphical analysis of the likelihood surface are studied in Chapter 6. The posterior distributions of the estimation model are derived in Chapter 7. Part III (Evaluating the concept) studies the small sample properties of the estimator. Part IV (Putting the concept to work) summarizes all the findings and poses questions for future research.