Data Envelopment Analysis: Balanced Benchmarking

ISBN-13: 978-1492974796
ISBN-10: 149297479X
List Price: US$ 100.00

Overview

The current book introduces the methodology of data envelopment analysis (DEA). DEA uses mathematical programming techniques and models to evaluate the performance of peer units (e.g., bank branches, hospitals and schools) in terms of multiple performance measures or metrics. These multiple performance measures are classified or coined as DEA inputs and DEA outputs. If the underlying process of DMUs is a production process, DEA can be used to examine the resources available to each unit and monitors the “conversion” of these resources (inputs) into the desired outputs.

Although DEA has a strong link to production theory in economics, the tool is also used for benchmarking in operations management, where a set of measures is selected to benchmark the performance of manufacturing and service operations. In the circumstance of benchmarking, the efficient DMUs, as defined by DEA, may not necessarily form a “production frontier”, but rather lead to a “best-practice frontier”. For example, if one benchmarks the performance of computers, it is natural to consider different features (screen size and resolution, memory size, process speed, hard disk size, and others). One would then have to classify these features into “inputs” and “outputs” in order to apply a proper DEA analysis. However, these features may not actually represent inputs and outputs at all, in the standard notion of production. In fact, if one examines the benchmarking literature, other terms, such as “indicators”, “outcomes”, and “metrics”, are used. The issue now becomes one of how to classify these performance measures into inputs and outputs, for use in DEA.

DEA’s empirical orientation and absence of a priori assumptions have resulted in its use in a number of studies involving efficient or best-practice frontier estimation in the nonprofit, regulated, and private sectors. DEA applications involve a wide range of contexts, such as education, health care, banking, armed forces, auditing, market research, retail outlets, organization effectiveness, transportation, public housing, and manufacturing. DEA is a balanced benchmarking tool that will help organizations to examine their assumptions about their productivity and performance.

The book provides students, researchers, and practitioners with a solid understanding of the methodology, its uses and potentials in business analytics.

Chapters 2-8 contain homework problems. The book uses Excel spreadsheets and VBA to automate the DEA calculations.

Table of Contents

Chapter 1: Introduction to Linear Programming
Chapter 2: Data Envelopment Analysis
Chapter 3: Efficiency Ratio and DEA Multiplier Models
Chapter 4: DEA Dual Models
Chapter 5: DEA Models and Returns to Scale
Chapter 6: DEA Models for Special Cases
Chapter 7: Slack-Based and Non-radial DEA Models
Chapter 8: Restricted Multipliers
Chapter 9: Super Efficiency
Chapter 10: Productivity Change
Chapter 11: Context Dependent DEA
Chapter 12: Flexible Measures
Chapter 13: Cross Efficiency
Chapter 14: Benchmarking Models

Download detailed Table of Contents (115 kb)

Preface

Employees who seem to work the least can often be the most productive. Business units that boast high profitability can sometimes be the least efficient.

The current book introduces the methodology of data envelopment analysis (DEA). DEA uses mathematical programming techniques and models to evaluate the performance of peer units (e.g., bank branches, hospitals and schools) in terms of multiple performance measures or metrics. These multiple performance measures are classified or coined as DEA inputs and DEA outputs. If the underlying process of DMUs is a production process, DEA can be used to examine the resources available to each unit and monitors the “conversion” of these resources (inputs) into the desired outputs. Although DEA has a strong link to production theory in economics, the tool is also used for benchmarking in operations management, where a set of measures is selected to benchmark the performance of manufacturing and service operations. In the circumstance of benchmarking, the efficient DMUs, as defined by DEA, may not necessarily form a “production frontier”, but rather lead to a “best-practice frontier”. For example, if one benchmarks the performance of computers, it is natural to consider different features (screen size and resolution, memory size, process speed, hard disk size, and others). One would then have to classify these features into “inputs” and “outputs” in order to apply a proper DEA analysis. However, these features may not actually represent inputs and outputs at all, in the standard notion of production. In fact, if one examines the benchmarking literature, other terms, such as “indicators”, “outcomes”, and “metrics”, are used. The issue now becomes one of how to classify these performance measures into inputs and outputs, for use in DEA.

Since DEA was first introduced in 1978, over 4,000 DEA-related articles have been published. Researchers in a number of fields have been quick to recognize that DEA is an excellent methodology for modeling operational processes and measuring productivity. As well, an important aspect of DEA is its productivity management capability. Specifically, because there are certain time series structures within the DEA suite of models (e.g., the Malmquist model and window analysis), these permit the management of performance from one period to the next.

DEA’s empirical orientation and absence of a priori assumptions have resulted in its use in a number of studies involving efficient or best-practice frontier estimation in the nonprofit, regulated, and private sectors. DEA applications involve a wide range of contexts, such as education, health care, banking, armed forces, auditing, market research, retail outlets, organization effectiveness, transportation, public housing, and manufacturing. DEA is a balanced benchmarking tool that will help organizations to examine their assumptions about their productivity and performance.

The book contains fourteen chapters as summarized below.
Chapter 1 introduces the ideas of Linear Programming (LP), a methodology central to the DEA tool.

Chapters 2-4 present the basic DEA models. These chapters show how DEA is developed and interpreted.

Chapter 5 introduces the concept of returns to scale in DEA, and the related basic DEA models.

Chapter 6 presents several DEA approaches for dealing with inputs and outputs that have their own special characteristics. For example, some inputs may be non-discretionary and are not under the control of DMU management. Some outputs can be undesirable, for example, waste production.

The models discussed in chapters 2-6 have either an input or output-orientation. Input-oriented models consider the possible (proportional) input reductions while maintaining the current levels of outputs. The output-oriented models consider the possible (proportional) output augmentations while keeping the inputs at their current levels. Going beyond these one-sided orientations, Chapter 7 introduces models that consider (i) both input reductions and output increases, and (ii) non-proportional input/output changes.

While DEA does not require a priori information on the tradeoffs or relative importance among the inputs and outputs, incomplete information on relationships among the factors may in fact be available. Chapter 8 examines how such information can be incorporated into the models, thus further refining the DEA results.

Chapter 9 presents the super-efficiency DEA models, where the unit under evaluation is excluded from the reference set. The super-efficiency DEA model is closely related to the approaches discussed in chapters 10 and 11.

Chapter 10 presents a DEA model (called the Malmquist productivity index model), for measuring productivity change over time. This structure makes an important link to statistical time series models.

Adding or deleting an inefficient DMU or a set of inefficient DMUs does not alter the efficiencies of the existing DMUs and the best-practice frontier. The inefficiency scores change only if the best-practice frontier is altered. i.e., the performance of DMUs depends only on the identified best-practice frontier. In contrast, researchers in consumer choice theory point out that consumer choice is often influenced by the context. e.g., a circle appears large when surrounded by small circles, and small when surrounded by larger ones. Considering this influence within the framework of DEA, one could ask “what is the relative attractiveness of a particular DMU when compared to others?” Chapter 11 presents a context-dependent DEA approach to measure the extent to which the relative attractiveness of DMUx compared to DMUy depends on the presence or absence of a third option, say DMUz (or a group of DMUs).

In DEA, it is assumed that the input versus output status of each performance measure related to the DMUs is known prior to the application of the model. However, sometimes a measure can be regarded as either (or both) an input and output. For example, in evaluating the performance of universities, graduate students can play the role of either an input (a resource available to faculty members, effecting their productivity), or as an output (trained personnel, hence a benefit resulting from research funding). Medical interns have a similar interpretation in the evaluation of hospital efficiency. n many problem situations such as those described, the input versus output status of certain measures can be deemed as flexible. Chapter 12 presents an approach aimed at facilitating the derivation of the input/output status of variables, when flexibility is an issue.

While DEA has been proven an effective approach in identifying the best practice frontiers, DEA can be viewed as a tool for self-evaluation. Chapter 13 presents DEA cross efficiency method which is developed as a DEA extension to rank DMUs with the main idea being to use DEA to do peer evaluation, rather than in pure self-evaluation mode.

Chapter 14 introduces a DEA-based benchmarking approach where one group of DMUs is compared to another group of DMUs.

We would like to thank Dr. Feng Yang for his help in developing the problems in chapters 2-8. We would also like to thank Christie Holmes, Brian Litke, and Christine Tang for developing the cases in this book. We thank Professor Hiroshi Morita for pointing out typos in the manuscript while translating this book into Japanese. However, any errors in the book are entirely our responsibility, and we would be grateful if any such errors can be brought to our attention.

Wade D. Cook and Joe Zhu, November 2013.

Japanese Version



Published by Shizuoka Scholarly Publishing
Release date: 2/28/2014
ISBN-10: 4864740305
ISBN-13: 978-4864740302

Translated by
Professor Hiroshi Morita
Osaka University

Previous version of the book