Comparing Environmental Risks

Tools for Setting Government Priorities

J. Clarence Davies

• Publisher: Routledge
• Year: 2014
• eBook ISBN: 9781135890612
• Topic: Biological Sciences
• Subtopic: Ecology
• Index: Biological Sciences

The budgetary squeeze of the 1990s has made it obvious that the government cannot address every possible environmental problem. Comparative risk assessment (CRA) is increasingly advanced as the means for setting realistic priorities. RFF's Center for Risk Management commissioned background papers from leading experts on CRA for a meeting with federal regulatory officials. Comparing Environmental Risks presents the revised papers of this workshop. Representing the state of the art on programmatic CRA, its methodological analyses and practical recommendations will be invaluable to government officials, independent analysts, and anyone studying environmental policy.

COMPARING

ENVIRONMENTAL RISKS

1

Comparative Risk Analysis in the 1990s: The State of the Art

J. Clarence Davies

Early in 1987 a small group of people met in the office of the administrator of the U.S. Environmental Protection Agency (EPA). The meeting had been convened to decide how to distribute a report that the agency had just completed. The discussion that day ranged broadly, including a suggestion that the report be given to Congress in a plain brown wrapper with no indication of authorship, because no report of this kind had ever been done by either EPA or any other government agency. The report, entitled Unfinished Business, was the first study to compare the risks addressed by the various EPA programs. Neither Administrator Lee M. Thomas nor anyone else in the room had a clear idea what the reaction to the report would be, but they chose to distribute the report because its findings were potentially important to the future of EPA and to the future course of environmental policy.

The reactions in the press and Congress to Unfinished Business were generally favorable, and various individuals concerned with environmental policy began thinking about the potential power of comparative risk analysis as a tool for sorting out and setting environmental priorities. Two years after the release of Unfinished Business, William K. Reilly, who had just been appointed EPA Administrator by President George Bush, decided to make risk-based priority setting a theme of his administration. That theme has been widely and heatedly debated since then, and it remains a central issue to much of federal environmental policy. As far as the contributors to this book are concerned, it is more a question of how best, not whether, to apply certain risk-based techniques to realize policy goals. Before discussing the perspectives of these writers, though, I would like to comment on how the risk paradigm in general, and comparative risk analysis in particular, evolved as a political issue in the mid-1990s, and how this book in its turn emerged as part of that development.

COMPARATIVE RISK ANALYSIS BECOMES A POLITICAL ISSUE

In the early 1990s several forces combined to put comparative risk analysis (CRA)—the relative ranking of risks—at the top of the environmental policy agenda. The budgetary squeeze at all levels of government made it more obvious than ever that not every environmental problem could be addressed—somehow priorities had to be set. Criticism of environmental programs in general and of EPA in particular focused in part on the low levels of risk posed by the problems being regulated. State and local governments chafed under federal requirements to spend significant amounts of money on problems that the states and localities considered both low risk and low priority. A number of local officials focused on CRA as a tool by which they could persuade Washington to pay attention to local priorities.

The state and local interest in CRA soon resonated in the U.S. Congress, which turned its collective attention for the first time to the complex and obscure topic of risk analysis. True, a few lonely congressional voices had previously focused on risk analysis. Congressman Don Ritter had introduced risk-based legislation as early as 1979: Senator Daniel Patrick Moynihan, since 1991, had been proposing legislation requiring EPA to periodically conduct a CRA of its programs. However, the 103rd Congress was the first to put risk analysis high on the agenda.

Because congressional understanding of risk analysis in general, and CRA in particular, was not very deep, it generally failed to distinguish either between individual risk assessments and comparative risk analyses, or between the two types of CRA: a large-scale comparison of programs (or problems) on the one hand and comparison of specific risks on the other. Thus the debates about risk incorporated business concerns about EPA’s methodology for performing individual risk assessments and a general concern about government overregulation with the interest in using risk analysis as a basis for setting program priorities.

For all its confusion of risk categories, the 103rd Congress did indeed focus legislatively on the topic with the Johnston Amendment, first proposed by Senator Bennett Johnston of Louisiana as an amendment to a bill elevating EPA to cabinet status. The Johnston Amendment would have required EPA to do a benefit-cost analysis of major regulations and to compare the risk being regulated to other risks. Controversy over the amendment was sufficient to prevent passage of any major environmental legislation. The Johnston language was added to a Department of Agriculture reorganization bill that was enacted, but the enacted language applied only to Agriculture regulations, and no other legislative action on risk was enacted.

After the 1994 elections, the newly dominant Republicans in the 104th Congress made risk-oriented legislation part of the “Contract with America” that they had pledged to uphold during the campaign. Within the first three months of the legislative session the House had passed a detailed regulatory reform bill with risk analysis as a major feature. (As of mid-October 1995, the Senate had not passed comparable legislation. Senator Robert Dole, the majority leader, had made three attempts to close off debate on the regulatory reform measure but had failed each time.) Both bills contained detailed requirements for risk and cost-benefit analyses to be conducted on all major regulations, and both expanded the scope of judicial review of the analyses.

The congressional interest in risk analysis had its executive branch analog. Risk assessments had been used routinely as a decisionmaking tool in EPA. However, programmatic CRA, which was pioneered by EPA, had not been done by any other agency and was not a routine procedure at all within EPA. The President’s Office of Science and Technology Policy (OSTP) had a strong interest in seeing CRA used more widely to set priorities.

Under the leadership of Mark Schaefer, OSTP turned to the Center for Risk Management at Resources for the Future (RFF) for assistance in formulating methods for broader use of CRA. From that effort, as described in the preface to this book, RFF organized first a preliminary meeting in August 1993 that pointed the way to further exploration into CRA methods. Funding was found to assist in this exploration, resulting in the February 1994 conference and its papers and, finally, this book.

The initial context of the conference for which the chapter material was prepared should be kept in mind when reading them: they were written primarily for a particular audience—federal regulatory agencies—and for a particular purpose—to encourage the use of CRA. However, because these writings represent the state-of-the-art of programmatic CRA, and because the importance and implications of that art reach far beyond the agencies, the contents of this volume will be valuable to a much wider audience.

THE SCOPE OF THIS BOOK

This book is structured as a logical progression, moving from a description of the CRA process to its history, to general principles, and then to the specifics of how to do it. Taken as a whole, the book outlines the evolution of CRA and its surrounding controversy, summarizes lessons learned from past implementation efforts, and suggests new ways of using CRA.While the individual chapter authors are not in total agreement with each other, they do agree to a remarkable degree both about the desirability of doing CRA and about the principles that should guide the conduct of CRAs.

In the book’s second chapter, I outline some key questions, the decision points that an agency must face when undertaking a CRA. Most of the questions are applicable to both types of CRA, but the emphasis is on programmatic CRA. I outline the options for answering the questions and in some cases make specific recommendations.

The chapter by Richard Minard presents the history of CRA, particularly within EPA, and details six of the state and local CRA efforts that EPA has sponsored. The essay reveals much about the methodology and the politics of CRA. It shows how a variety of governments have wrestled with the issues described in the first essay. More state-level CRAs have been done than any other type of large-scale CRA, and Minard’s essay distills this experience.

Frederick Anderson’s chapter describes the background factors that set the context for the congressional attention to risk. He then describes the positions that the principal stakeholders—environmentalists, industry, state and local governments, and the Clinton administration—have taken on risk assessment and CRA. Finally, Anderson draws a series of lessons and recommendations aimed at improving the use of CRA in the executive branch. He outlines how the administration might promote receptivity to risk-based priority setting while responding to criticisms and concerns.

The chapter by John Graham and James Hammitt makes a series of suggestions to refine the framework for conducting CRAs in the federal government. As they state, “This chapter is intended to provide some conceptual guidance to administration and congressional officials who are engaged in the task of promoting a new priority-setting process informed by comparative risk.” Graham and Hammitt particularly stress the importance of ranking risk-reduction options as well as baseline risks—in other words, of looking at the incremental risks that would be averted by specific actions.

The final chapter, written by a group from Carnegie Mellon University (M. Granger Morgan, Baruch Fischhoff, Lester Lave, and Paul Fischbeck) proposes a procedure that agencies can adopt to rank systematically the risks for which they are responsible. Whereas the other authors suggest principles, caveats, and questions, the Carnegie Mellon team says in effect: “If you want to do a CRA, here’s how.” The methodology it describes is comprehensive and incorporates a number of methodological innovations, such as the use of public panels. The RFFOSTP workshop spent more time on the Carnegie Mellon paper than on any other because it most clearly spoke to the question of what specific steps an agency would take to do a CRA.

A BASIC VOCABULARY OF RISK

A multiauthored volume about any discipline undergoing transformation inevitably raises problems of terminology. CRA is just such an energetically developing field, so this book is no exception. For quite legitimate reasons, the precise definition and usage of some terms vary considerably among risk professionals, so little effort was made to impose a uniform set of definitions on the chapter authors. Therefore, the same term might have a slightly different meaning and, conversely, the same phenomenon might be referred to by different names. The following paragraphs identify and discuss the basic risk terms that most frequently may cause confusion. These meanings derive from discussions with Paul Portney, and are based in turn on our understanding of the common usage of these terms and their logical interrelations.

Basic Risk and the Four Pillars of Risk Analysis

To start, risk itself is simply the likelihood that injury or damage is or can be caused by a substance, technology, or activity. Often a risk estimate contains an explicit probability factor: “There is a 50% chance that the structure will collapse and cause the death of three people.”

In similar fashion, when describing the risk-based disciplines, risk analysis is the most general term, which encompasses comparative risk analysis, risk assessment, risk management, and risk communication.

Comparative Risk Analysis (Risk Ranking). Comparative risk analysis—CRA—is also termed risk ranking or relative risk ranking. One type of CRA consists of comparing two relatively well-defined types of risk. It could consist of comparing risks from two similar sources—the cancer risk from two different pesticides—or it could compare two dissimilar risks—the risk of dying in a canoeing accident with the risk of dying from being exposed to benzene. A CRA could also be a comparison of the risk from two different types of controls on the same source, such as comparing the use of chlorine to ozone for purifying water.

A second type—programmatic CRA, the focus of this volume—is used for setting regulatory and/or budgetary priorities and involves comparison among a large number of risks. This larger-scale CRA also differs from the first type by involving a great deal more value judgment, and is as much a philosophical as a scientific effort. Examples of programmatic CRA would be establishing the relative importance of various pollution control programs on the basis of risk reductions or establishing priorities based on risk for cleanup among a group of hazardous waste sites.

Risk Assessment. Risk assessment is a set of analytical techniques for answering the question: How much damage or injury can be expected as a result of some event? Although risk assessment was first developed to estimate the probabilities of an accident in a particular type of technology (a nuclear reactor, for example), risk assessment methodology has been most elaborately developed for estimating the cancer effects from chemicals based on laboratory testing of rodents. The general term “risk assessment” is often mistakenly equated with this particular type of cancer risk assessment.

The most generally accepted formulation of risk assessment is the following four-step process, devised by a committee of the National Academy of Sciences (NAS 1983).

1. Hazard identification, which identifies the type of injury that can be caused (for instance, chemical X can cause liver damage).

2. Dose-response assessment, which estimates the relationship between exposure to a harmful substance (or event) and the resultant harm (for instance, exposure of X parts per million of substance Y for a period of 2 hours can produce liver damage).

3. Exposure assessment, which estimates how much of a substance will reach a target population or how much of a population will receive some exposure to a substance.

4. Risk characterization, which combines the information from steps 2 and 3 to estimate the amount of injury or damage that will be caused by a substance, technology, or other risk source.

Risk assessments can be done for acute health effects such as workplace injuries, chronic health effects such as cancer or birth defects, and ecological effects such as reductions in species or damage to trees. All risk assessments suffer from a fairly high degree of uncertainty. The statistically calculated uncertainty associated with most chronic health risk assessments is several orders of magnitude (there is a hundredfold or thousandfold spread between the high and low plausible estimates). The real uncertainty is even larger because of controversy about the basic premises of most risk assessments, such as the validity of animal testing as a way of predicting human health risks.

Risk Management. Risk management developed as a contrasting term to “risk assessment.” Risk management considers the social, economic, and political factors involved in risk analysis, determining both the acceptability of damage that could result from an event or exposure and what, if any, action should be taken with regard to the risk of that damage. Compare this to risk assessment, which helps estimate the likelihood that such-and-such damage could result from an event or exposure.

In some instances, risk management refers to all risk-related policy processes other than risk assessment—agenda setting, decisionmaking, implementation, evaluation—whereas in other situations this term is limited to decisionmaking about risk.

Risk Communication. Risk communication is just that: conveying information about risk. Such communication can range from simple warning labels to product data sheets to hazardous site databases to public hearings. Originally, risk communication was generally one-way, with the risk experts (whether technical advisor or risk manager) trying to impart expert knowledge to the public. The view that the process needs to be considered as a mutually informing interchange among the interested parties—including the public—is becoming more widely accepted.

A STRONG FOUNDATION OR A...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Table of Contents
• Foreword
• Preface
• 1 Comparative Risk Analysis in the 1990s: The State of the Art
• 2 Ranking Risks: Some Key Choices
• 3 CRA and the States: History, Politics, and Results
• 4 CRA and Its Stakeholders: Advice to the Executive Office
• 5 Refining the CRA Framework
• 6 A Proposal for Ranking Risk within Federal Agencies
• Index