OPERATIONS RESEARCH AND THE PUBLIC SECTOR
Edited by S.M. Pollock, M.H. Rothkopf and A. Barnett
This is also why the level of treatment of the topics within this volume may appear to vary widely - a natural phenomenon when one considers that each reader is bound to bring to his or her perusal an equally wide variation of backgrounds, education and technical skills.
The first chapter, by Pollock and Maltz, sets the stage by discussing the suggested boundaries for what we mean by the 'public sector', 'operations research' and 'management science'. It also presents a brief history of the applications of the latter two to the former.
In Chapter 2, Gass provides a different kind of introduction, in which he discusses the nature of an 'application' in the public sector and how it differs from one in the private sector. He examines ways that public sector applications can be improved and presents the ethical issues that arise in the context of model use for public sector advocacy.
"Everyone concedes that models can fail, but many modelers appear furtively to believe that their own efforts are invulnerable to that fate." So says Barnett in his 'admonitional' Chapter 3, in which he also contends that the creativity so often apparent in the construction of models is less evident when the need arises to validate them. This serious omission can have particularly serious consequences when dealing with issues involving human life and welfare. Some readers will be surprised at his re-telling of the argument that the R2 measure, so pervasive in regression analysis, may be the most overrated statistic in the history of mathematics.
It is well established that the first organized use of operations research as a 'discipline' was in the analysis of military operations. No one doubts that analytical capabilities continue to be vital in both tactical and strategic aspects warfare. As Washburn makes clear in Chapter 4, it continues to be an important, if not infallible, contributor to national defense. He also presents arguments as to why military planners are well advised to use these capabilities to ferret out the implications of imperfect (and thus often probabilistic) information.
Operations research has especially strong links to the transportation field; it is no accident that Transportation Science was the first of ORSMs specialized journals. Beyond solving myriad problems in scheduling and routing by land, sea, and air, OR can help in such unexpected areas as designing airport terminal buildings. Odoni, Rousseau and Wilson offer, in Chapter 5, a vivid spectacle of OR-in-motion.
Problems in planning for and operating emergency services (e.g. police, fire, and ambulances) have proved highly amenable to such techniques as queuing analysis and mathematical programming. OR studies have generated a clearer understanding about the allocation of critical resources to such services. They have also made the case for such unsettling decisions as delaying the response to a current fire alarm to prevent a far greater delay for 'the fire next time'. In Chapter 6, Swersey surveys developments in this area in which OR models have doubtless influenced many lives.
In Chapter 7 on Crime and Justice, Maltz points out that Poisson derived his seminal probability distribution - practically the 'union label' of operations research - during the course of modeling jury behavior. More recent OR activities in this area have focused on the stochastic processes that underlie data about the commission and punishment of crime. This work has deepened and sometimes radically transformed the public understanding of criminal activity, leading to the observation that operations researchers have had an influence on criminology far exceeding the proportion of their numbers.
The energy industry has seen extensive private sector application of operations research modeling. In Chapter 8, Weyant, the director of Stanford's Energy Modeling Forum, describes the use of models to analyze energy policy choices. These models, spanning the range from engineering to economics, are crucial to understanding the impact of various alternative energy generation and distribution strategies on the general public.
In bitter forest-usage debates in the Pacific Northwest, it has sometimes seemed unclear whether the endangered species is the spotted owl or the lumberjack. Operations research cannot provide moral guidance in such matters but, as Golden and Wasil explain in Chapter 9, it can explicate tradeoffs between particular policy choices. In a parallel fashion, they also present the application of appropriate models to the understanding of impacts of various actions on the availability and usage of other natural resources: fisheries, wildlife and water.
In Chapter 10, ReVelle and Ellis consider a different but related concern: modelling the effects of management decisions on air and water quality. Different methods of generating energy, producing goods or extracting raw materials cost differing amounts and yield different levels of air and water pollution. Understanding the appropriate balance of such activities naturally produces problems in constrained optimization, one of the classical tools of OR.
Nobody wants a garbage dump located near his or her home, but no one wants to live in a society that has no garbage dumps. Meindorfer and Kunreuther deal with conundrums of this kind in Chapter 11, where they present the analyses behind rational approaches to siting hazardous or other noxious facilities. They describe the important features of this 'facilities siting' problem, discuss approaches for improving structuring of the problem, and review research devoted to siting (and related route selection) problems.
The Delaney Amendment made everything seem easy when it passed the U.S. Congress: a substance would be banned if it had even the slightest carcinogenic potential. Taken literally, the amendment would indeed have eliminated cancer because we would have starved to death in a few weeks. But defining, estimating and managing risk is a more subtle business, as Lave makes clear in a wide-ranging Chapter 12.
Pierskalla and Brailer's Chapter 13, on the use of operations research in analyzing health care delivery, is particularly timely now that the United States is embarked upon a national debate over changing the way health care is delivered. Dealing with death and physical impairment as possible decision outcomes, within the constructs of formal operations research models, is a challenge that can be met in dealing with health care analysis. The discussion evaluates these methods, and provides extensive references to an extensive body of relevant literature.
The analysis of sports and athletic competitions may not be considered to fit easily within the rubric of understanding a 'public sector' activity. It does share at least one factor in common, however: the use of a financial 'bottom line' is often an inadequate way of characterizing the more interesting problems. Moreover, anything that contributes to the fairness and excitement of sporting activities has far-reaching economic and social implications for the public sector. (The Olympics are watched by half the people in the world.) Gerchak reminds us in Chapter 14 that operations research is by no means absent from the sporting arenas, and may have had more influence (or at least may deserve more) than is commonly recognized.
Suppose the Census reveals that, over the previous decade, your state's population has grown while all other states had no change in their populations. A reapportionment method, supposedly rooted in the 'one person, one vote' principle, could wind up reducing your state's representation in Congress! In Chapter 15, Balinski and Young subject various apportionment rules to deep scrutiny, identifying the axioms implicit in each system and suggesting anomalies to which they can lead.
In Chapter 16, Anderson addresses the ubiquitous public-sector concern about how to reach decisions among many parties who have conflicting preferences. The obvious application is to voting as a means of electing public officers, but generalizations to other group selection processes are also covered. In Chapter 17 he presents the principles by which one can understand ranking and pairwise comparisons, analyses that encompass a wide spectrum of activities, ranging from weapon system evaluation to college football and basketball polls.
The last two chapters deal with methodological issues (in contrast to the previous chapters' arena-based materials) of particular concern in the public sector.
In analyzing public sector problems, management scientists and operations research analysts are immediately faced with the task of defining scales of measurement related to the phenomena they are studying. In Chapter 18, Roberts spells out the issues that arise when analysts create such measurement scales and indicates the kinds of difficulties that can arise if these scales are not chosen appropriately.
In seeking a politically acceptable way to purchase property for the government, or to transfer government rights, governments often turn to the use of auctions. In the final Chapter 19, Rothkopf surveys critically the wide variety of mathematical models of auctions, and analyses their use in the design of competitive systems.
Initiating and organizing the material in this volume, and dealing with the authors over the past few years, has been a challenging and rewarding experience for us. We hope the reader will find the breadth and depth of the applications and methods presented here as exciting and provocative as we did. With some irony we recognize that, due to the obvious success in applying operations research methods and approaches to the 'public sector', there is a good possibility that within the next few decades there will be little need to single them out as anything but accepted 'good practice' in this important arena.
Stephen M. Pollock
Arnold Barnett
Michael H. Rothkopf
Complete chapters on ScienceDirect
[Description and order information]
Home | Elsevier Sites | Privacy Policy | Terms and Conditions | Feedback | Site Map | A Reed Elsevier Company