Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 3
1
Introduction
Regulatory agencies regularly confront difficult trade-offs among im-
portant values, such as health, safety, longevity, and the monetary and
nonmonetary costs and benefits associated with protecting human health
and well-being against the hazards of modern life. Those responsible for
reducing deaths and illnesses due to encounters (sometimes years earlier)
with toxic substances and other potential health and safety hazards must
often assess uncertain events. They must also balance benefits and costs
that are difficult to characterize or measure. Individual analysts and entire
regulatory agencies, such as the Environmental Protection Agency (EPA),
have devoted considerable effort to improving benefit-cost analysis, cost-
effectiveness analysis, and other ways of characterizing trade-offs among
key values. However, some analytic approaches and techniques and the
details of their application remain controversial.
For example, once scientists have identified the risks involved in prob-
lems such as pollutants in the air and water, unsafe factory conditions,
dangerous consumer products, and threats from toxic chemicals, regulatory
attention shifts to the question of what kind and how many resources should
be devoted to reducing or avoiding the risk in question. Consider some
common examples: Which policies will do the most to improve the safety
of children's toys? Who will pay for the health effects of asbestos exposure?
Should we require copper smelters to reduce or eliminate fugitive arsenic
emissions? How do we know whether preventing further ozone depletion
is worth a particular effort? Such choices, because they involve inherently
scarce resources, attach values to the consequences of the actions involved.
This volume reports on the results of a conference planned and con-
ducted by an independent committee of volunteers appointed by the Na-
tional Research Council to address an important set of issues encountered
in using formal approaches to valuing risks, costs, and benefits in health
and safety policy making. During the 1980s, interest in comparing benefits
and costs of proposed regulations was stimulated in EPA and several other
3
OCR for page 4
4
INTRODUCTION
federal regulatory agencies, partly in response to a sharper focus by the
President and the Office of Management and Budget (OMB) on reducing
private-sector expenditures associated with regulation. The difficulty of de-
veloping explicit values for human lives extended, disabilities avoided, and
others has led these agencies to search for practical yet theoretically defen-
sible approaches to risk valuation and to benefit-cost or cost-effectiveness
analysis. As a part of that effort, the committee organized a conference
that surveyed recent scholarship in law, philosophy, political science, and
economics and discussed what research, if any, would be likely to lead to
improved regulatory decision making, particularly benefit-cost analysis.
Neither the conference nor the more general search for improved
benefit-cost analysis has produced complete agreement within agencies,
across agencies, in Congress, in OMB, and among interest groups about
valid ways to improve valuation of risks, costs, or benefits. Rather, the
search has revealed, and in some instances sharpened, disagreements about
the appropriateness of specific attempts to value risks for health, safety,
and environmental regulation and about methods and approaches for do-
ing so. Instead of integrating analysis of environmental, health, and safety
risks into complete discussions of regulatory decisions, those who oppose
or support particular regulations have often chosen to separate them from
consideration of other factors, to politicize the analysis, or both. And dis-
agreement about some of the ethical issues underlying benefit-cost analysis
can allow such politicization to continue. Consequently, agencies producing
analyses find themselves repeatedly challenged in court and in Congress for
exceeding their statutory charters, failing to account for important noneco-
nomic concerns-such as the special nature of human life-or inadequately
addressing other issues, such as effects that would be felt only a very long
time in the future.
The central policy-making problem for health and safety regulation is
thus the very practical question that assisted the committee in organizing
the conference, as well as its conclusions and recommendations: What is
gained by explicit treatment of the inherent values and trade-offs? Since
any regulatory decision involves resource expenditures by individuals, firms,
government, or other parts of society, some examination of resource expen-
ditures and what is received for them seems appropriate. But how should
this be done, especially when intangibles like human health are involved? Is
there any prospect for improvements in benefit-cost analysis that would pro-
duce increased consensus on its underlying assumptions and application?
Interest has grown among regulators, scholars, and analysts in exploring
some of the most troublesome issues underlying valuation techniques and
approaches in an effort to understand the sources of disagreement and the
possibility for achieving more consistent, valid, and practical approaches.
These issues include:
OCR for page 5
INTRODUCTION
(1)
(2)
s
administrative, legal, and statutory constraints on application of
benefit-cost analysis techniques;
the philosophical foundations of using a common monetary met-
ric to represent human life and health;
moral issues involved in comparing health-risk benefits and costs
across time and distance, such as individual and collective prefer-
ences for who should bear the burdens or benefits of risks across
time and distance;
the adequacy of scientific information as a basis for placing values
on health risks in particular instances; and
effects on regulatory policy of new research knowledge about
decision making.
The balance of this introductory chapter further explores questions
raised for researchers, regulatory policy makers, and agency analysts by the
use of the valuation techniques we have labeled benefit-cost analysis.
RISK ASSESSMENT ANI) BENEFIT-COST ANALYSIS
In addressing environmental, health, and safety problems, regulatory
agencies are often asked to develop an understanding of both risks and
resources. They are asked to assess the risks associated with specific
health and safety hazards. They may also analyze the costs and benefits
of alternative ways of reducing those risks. Although they operate under
different statutes (in some cases, several for one agency), EPA and other
regulators such as the Occupational Safety and Health Administration,
the Food and Drug Administration, and the Consumer Product Safety
Commission have been required for more than a decade to formally
assess the health risks associated with exposure to toxic substances and
other potential hazards, including air and water pollution, food additives
and contaminants, and products and processes in the workplace. During
that time there have been a number of efforts to improve the physical,
chemical, and biological bases for policy making. More recently, there
have been several efforts to examine the process for assessing risks at
the Environmental Protection Agency and other federal health and safety
regulators.
One such effort, by a committee of the National Research Council,
examined the process by which federal agencies characterize and determine
the risks of cancer for policy-making purposes. The committee's report, Risk
Assessment in the Federal Government (National Research Council, 1983),
examined procedures used by agencies in the early 1980s to determine
cancer risks. It proposed that development of scientific information about
specific hazardous risks be separated analytically from decision-making and
management activities of a regulatory agency. Such a separation would
OCR for page 6
6
(2) DOSE-RESPONSE
ASSESSM ENT
(5) CONTROL
ASSESSMENT
Risk Assessment
(1 ~ HAZARD
, I DENTI FICATION
(4) RISK
CHARACTERIZATION
\ /
(7) REGULATORY DECISION
Risk Management
FIGURE 1 Risk assessment and management process.
INTRODUCTION
(3) EXPOSURE
ANALYSIS
(6) BENEFIT-COST
ANALYSIS
(REGULATORY IMPACT
ASSESSM ENT)
allow for scientific criteria to be the primary criteria in guiding research
and data gathering.
The report also constructed a framework for organizing diverse kinds
of hazard-related information needed for a complete program of risk man-
agement; it is summarized and adapted in Figure 1. In this scheme an
agency might begin with a careful process of risk assessment, including
data gathering, scientific testing, and evaluation of results, to learn more
about human and animal responses to various doses of a potential hazard
and about what exposures to that hazard people may be experiencing.
The 1983 report acknowledged the reality of contemporary risk assess-
ment by identifying areas in which gaps in scientific data and theory often
appear. It proposed that agencies develop guidelines for using conservative,
scientifically appropriate inferences to bridge the gaps for policy-making
purposes rather than either taking ad hoc inferential leaps or postponing
action until all scientific uncertainty is removed. 1b properly identify the
strengths and weaknesses of any assessment, the report emphasized the
need for assessors to make assumptions explicit as well as the need for peer
OCR for page 7
INTRODUCTION
7
reviews of the scientific process and results of assessments. In response
to proposals such as these, federal regulators have invested considerable
effort in recent years to improving assessments of hazards in order to better
characterize the risks involved.
Although the 1983 report focused primarily on risk assessment, several
of its conclusions and recommendations are also relevant to the analysis
of benefits and costs in support of risk management. The term benept-cost
anahysis refers to a family of approaches and techniques for expressing
alternative ways of achieving an objective in quantitative comparable terms.
For example, the classical approach is to express all benefits and costs
associated with each alternative in dollars as a function of time, discount
the future benefits and costs at some appropriate rate, and then compare
the alternatives on the basis of the current value of the net benefits or on
the basis of an internal rate of return (Quade, 1982:58, uses this common
definition; among many others, see Kneese, 1989:295-298~.
This classical approach most closely resembles an ideal definition that
cannot be applied in many instances due to practical limitations. More
often, agencies use a version of benefit-cost analysis that focuses on the
cost-effectiveness of alternative policies. That is, a single benefit (e.g., a
reduction in pollution) might be analyzed against several potential options
with various costs. The distinction is important in that costs but not benefits
are being compared. Nevertheless, this volume focuses on philosophical
and other assumptions underlying comparisons inherent in analysis of costs,
benefits, or both. Therefore, following common practice, in this volume
except as otherwise specified, the term benefit-cost amasses also refers to
cost-effectiveness analysis and other variations on this theme. For example,
after evaluating the potential health risks of a pollutant or toxic substance
or consumer product design, regulators may examine the costs of alternative
strategies for a uniform reduction in the risk (e.g., eliminate the offending
item, mitigate its effects, redesign it, or do nothing), and then compare
explicitly the costs of those alternatives. Information about risks, costs, and
(when permitted and/or possible) benefits may be summarized and used by
regulatory policy makers as part of the basis of a regulatory decision or in
order to support one.
REGULATORY PRACTICE
Regulatory practice has been far less consistent than such a scheme
would indicate, however. Approaches to valuing health and safer risks
have varied considerably by issue, between agencies, even within agencies,
and over time. The sources of such differences are numerous. First, with
certain exceptions, Congress grants each regulator considerable discretion
in carrying out broad legislative mandates. Regulatory agencies are allowed,
OCR for page 8
8
INTRODUCTION
encouraged, and even required to exercise discretion in implementing
statutes. Second, statutes (and case law) vary as well, some specifying that
regulators will consider costs and benefits of alternative policies, others
actively prohibiting such economic considerations. While most agencies
have considered benefit-cost issues in regulatory matters, statutory variation
allows some agencies to pursue the latest techniques and approaches more
vigorously than others. Third, and possibly most relevant to the current
attention being given to regulatory risk valuation issues, is the evolution
of presidential interest in and ability to affect the application of cost
calculations in regulatory policy making. Efforts by OMB to guide agency
use of statutory discretion to promote greater analytic rigor and consistency
in federal approaches to regulatory risk valuation have sparked serious
debates about the prospects for reaching those goals.
The use of benefit-cost analysis In the federal government is long-
standing and well documented. Benefit-cost analysis has antecedents in
the early 1930s, as a way of evaluating public water projects. It emerged
on a large scale in the early 1960s in the Department of Defense and in
federal recreation programs with the use of the Planning, Programming,
Budgeting System (PPBS) and its successors. Application of benefit-cost
techniques in health and safety matters grew from several White House
policy review mechanisms. As an antecedent to formal benefit-cost analysis,
in 1971 OMB ordered EPA to submit significant regulatory proposals
to an interagency committee for nonbinding discussion. Subsequently,
President Ford's 1974 executive order required agency heads appointed
by the President to submit for OMB review Inflation Impact Statements
(later entitled Economic Impact Statements) on proposed regulations likely
to have substantial economic effects. This requirement was used to prod
agencies to explain more explicitly the likely economic effects of regulations,
but no consistent agreement emerged as to how these statements would or
would not be used to inform regulatory decisions.
A Carter administration executive order issued in 1978 created a
new interagency forum, the Regulatory Analysis Review Group (RARG),
chaired by the head of the Council of Economic Advisers, to discuss
economic analyses (Regulatory Analyses) of 10-20 key proposed regulations
each year. Although benefit-cost analysis was not specifically part of the
RARG charter, agencies were asked to include "an analysis of the economic
consequences of each . . . alternative." In addition, agencies were directed
to solicit public participation, to choose the least burdensome alternative,
and to justify the choice. The RARG then reviewed, discussed, and made
public nonbinding comments on these analyses and proposed regulations.
Within a month after taking office, the incoming Reagan administra-
tion issued Executive Order 12291 (February 17, 1981), which replaced the
OCR for page 9
INTRODUCTION
9
RARG process with OMB's formal review and clearance of proposed reg-
ulations and which required agencies to prepare and to base all major new
regulations on formal benefit-cost analyses. The White House emphasized
that, within an agency's statutory constraints, the principles guiding regula-
tory decisions should be twofold: first, that regulation should be used only
in instances in which it is superior to market forces and, second, that reg-
ulation should be demonstrated to maximize net benefits to society. More
specifically, notable provisions of E.O. 12291 included requirements that
regulatory actions should not be undertaken unless potential benefits to
society from the regulation outweigh the potential costs to society; among
alternative approaches, the alternative involving least net cost to society
should be chosen; and that agencies set regulatory priorities to maximize
aggregate net benefits, taking into account the condition of particular in-
dustries involved. These requirements provided the rationale for OMB's
effort to increase the use of formal techniques for determining net benefits
to society. In the case of environmental, health, and safety regulation, the
executive order prompted OMB to require further efforts on the part of
agencies to value alternative policies for prevention of deaths, injuries, and
illnesses due to environmental and other hazards.
Overall, during the eight-year Reagan administration, OMB reviewed
approximately 30 regulatory impact analyses done in support of major new
proposed regulations. Between 1981 and 1985, 15 analyses were submitted
by EPA in the areas of environmental health and safety regulation. In
addition, during this same period of time, OMB chose not to review dozens
of other benefit-cost and cost-effectiveness analyses performed by EPA and
other regulatory agencies. OMB may have been limited in its coverage
of analyses by the size of its own analytical staff. Of approximately 35-40
professional staff in OMB's Office of Information and Regulatory Affairs,
nearly all were devoted to regulatory matters in the years l9Bl-1989.
Nevertheless, the number of regulations issued by EPA alone in that period
totaled over 2,000.
CONTINUING ISSUES
In actual regulation of health and safety problems, however, the regula-
tory agencies' use of formal approaches to valuing risks, costs, and benefits
remains unsettled. Particular issues associated with formal valuation remain
controversial and the subject of continuing debate. For example, there is
no single approach or set of methods for benefit-cost analysis now in use in
any federal regulatory agency with responsibility for health and safety. At
EPA, for example, of 15 Regulatory Impact Analyses submitted to OMB for
review between 1981 and 1985, 6 included direct comparisons of benefits
and costs of regulatory alternatives, and 9 concentrated primarily on costs
OCR for page 10
10
INTRODUCTION
of alternative actions but not benefits (EPA, 1987~. 1b become accepted,
even within one agency, any general approach or set of methods would
have to address and settle a continuing set of controversies concerning
methodological, conceptual, and moral challenges in characterizing and
valuing environmental and other health and safety risks. It was this situa-
tion that led EPA to request formation of the committee that prepared this
conference report.
In his keynote address to the conference (Chapter 2 in this volume),
Milton Russell, as former Assistant Administrator for Policy, Planning, and
Evaluation at the Environmental Protection Agency, points out graphically
the importance of these questions to the day-to-day decisions of regulatory
agencies and just how difficult it is to resolve the issues they raise. ~
illustrate the point, he asks the question "Where should the sludge from
municipal sewage plants be disposed?" By pointing out that every option
includes at least some risk to human health and reminding us that the
sludge has to go somewhere, he demonstrates how difficult it is to reach
satisfactory answers. Despite the difficulties and the imponderables, Russell
takes the position that it is irresponsible for a public official to decide
without tracing the consequences to the extent possible. In real life, public
officials face messy, incomplete answers. Russell feels that the legitimacy
of their decisions has to derive from the acceptance of those decisions by
the people affected, and that the provision of information about the likely
consequences is the basis for principled accountability.
Contextual and Legal Constraints
Benefit-cost analysis, like any set of decision tools, is affected by the
characteristics of the setting in which it is applied, including the specific
statutory requirements and restrictions, the distribution of policy-making
responsibilities, and the characteristics and behavior of a multiplicity of
interest groups. In order to gauge the possibility for developing more con-
sistent approaches to risk valuation, it is important to understand variations
among agencies in terms of such constraints.
What are the scientific, administrative, and political sources of
support for considering economic efficiency as an integral part of
risk-management decisions, and what are the constraints on doing
so?
· How do statutes, case law, presidential directives, congressional
pronouncements, and administrative practice provide guidance in
the development and application of benefit-cost analysis for regu-
latory decisions?
What are the key variations in the use of risk-control analysis by
federal health and safety regulators?
OCR for page 11
INTRODUCTION
11
In "The Politics of Benefit-Cost Analysis" (Chapter 3), R. Shep Melnick
focuses on the political and administrative forces that block wider imple-
mentation of formal benefit-cost analysis for federal regulatory decisions.
He outlines the sources of opposition to the use of analysis, as well as
the mechanisms that are used by major actors-Congress, public interest
groups, OMB, and the agencies-to shape general approaches and specific
risk valuation practices. Melnick raises the question of whether the current
legal and political context will allow for progress in integrating technical
benefit-cost analysis with regulatory decision making on health and safety
issues. A lack of progress may be due largely to controversies among key
policy makers on the moral validity of issues such as placing a value on
human life for regulatory purposes.
Approaches to Analysis
Some analysts and decision makers may find it attractive to value risks,
costs, and benefits in terms of a single arithmetically manipulable yardstick
that measures all relevant attributes, but various theoretical treatments of
value raise problems that must be dealt with in practical applications of the
approach. There is, for example, considerable debate about appropriate
ways to make comparisons across time and among illnesses, disabilities,
and economic or social groups.
What might be the foundation for a metric covering a variety of
health and safety effects?
· Can preference-based economic guidelines be accommodated with
political, moral, and other statements, especially with respect to
assigning, directly or indirectly, values to human life and health?
· What measurement problems are posed in valuing health risks (e.g.,
are there appropriate measurement scales, are health effects well
ordered)?
In "Benefit-Cost Analysis as a Source of Information About Welfare"
(Chapter 4), Peter Railton raises some of the doubts he has as a philoso-
pher about the approach. He addresses these doubts by asking the question
"What do I have to believe in order to do [benefit-cost analysis]?" In look-
ing for answers to this question, he considers whether such beliefs seem
tenable. For example, he inquires into whether theorists and practitioners
of benefit-cost analysis recognize limitations or treat it as sufficient for
policy determinations. Railton makes similar inquiries about the assump-
tions required to believe the scientific measurements on which benefit-cost
analyses are necessarily based and about assumptions associated with the
blanket application of discounting.
OCR for page 12
12
IN17?ODUCTION
Douglas E. MacLean, in "Comparing Values in Environmental Policies:
Moral Issues and Moral Arguments" (Chapter 5), addresses a related
set of questions. He reviews briefly several approaches to moral inquiry
common in the discipline of philosophy and describes how they might be
applied to valuation of health effects in environmental policy decisions.
He describes weaknesses in both the "low road" of expressed preferences
and the "high road" of theoretical or doctrinal moral reason. Examining
central paradoxes in the use of discount rates, especially with regard to
the valuation of human life, he points out necessary inconsistencies in
the universal application of analytic techniques for making environmental
policy decisions and proposes case-specific application of such techniques.
In particular, he raises problems applicable in any discussion in which
not all costs of a policy are borne by the same individuals or groups as
those who reap the benefits. These problems are relevant to intertemporal,
interpersonal, geographic, and interillness comparisons of risk.
How Much Information?
It is also important to understand how decision makers cope with
information that is imperfect and changing over time, especially since
analytic processes and the structuring of data affect the selection of what is
included in analyses. This raises a whole series of questions concerning the
timing of decisions, their interaction with information-providing activities,
and the aims and purposes of the organizations in which they are made.
.
Can procedures for valuing health risks be made politically feasi-
ble or acceptable; for example, can we agree on mechanisms for
deciding who should have a say in determining values and specific
regulatory policies?
How do beliefs about the major consequences of various decisions
affect our search for illuminating information?
· What is the relationship between the need for credible valuation
and the need for credible decisions?
In `'Environmental Policy Making: Act Now or Wait for More Infor-
mation?" (Chapter 6), Jeffrey E. Harris points out that regulatory decisions
about environmental hazards are routinely made in the face of huge un-
certainties. He characterizes the central problem in the dynamics of such
decisions as that of timing to act now or hold out for more information.
Harris suggests that a principal reason for waiting is the assumption that
environmental policies are irreversible. But he observes that in many cases
this involves a kind of self-fulfilling prophesy the longer we wait, the fewer
options remain and the more irreversible the decision appears. Harris also
suggests that waiting for more research may be equally misguided. The
OCR for page 13
INTRODUCT ON
13
best way to gather information, he claims, may most often be to implement
the policies and then to evaluate carefully their consequences. This strat-
egy, of course, assumes that the expected negative consequences of initial
implementation are small relative to the consequences of doing nothing.
Handling Uncertainty
Behavioral scientists have begun to challenge the simple theoretical
model of rational choice underlying most risl`-control analysis, suggesting
ways the model might be modified and applied in practical analyses. Much
of this work has focused on the handling of uncertainty in formal decision
approaches.
· How should analysts and decision makers acknowledge and cope
analytically with uncertainties in the science underlying decisions?
Are we sure enough of our ability to characterize the risks of
potential health hazards to use those characterizations as a basis
for valuation analysis?
How do we get better estimates of uncertainty for purposes of
practical decision making?
In "Choice Under Uncertainty: Problems Solved and Unsolved"
(Chapter 7), Mark J. Machina reviews the expected utility model and
describes several problems that beset it, including nonlinearity in probabil-
ities, preference reversal, and framing effects. He examines the empirical
evidence regarding each and suggests how these findings have, are likely to,
or should change the way economists view and model private and public
decisions under uncertainty. On balance, Machina feels the modern models
used to characterize behavioral deviation from expected utility theory are
completely consistent with well-defined individual preference ordering, and
hence with traditional welfare analysis. What to do about framing effects,
however, is itself a public policy issue, in Machina's view. We may be able
to look to other previous treatments of the presentation of information for
guidance.
CONCLUSION
Initial drafts of the papers in this volume served as provocative stimu-
lants for discussion at the conference. Following the papers is a concluding
chapter summarizing discussion among conference participants and the
steering committee's conclusions regarding the possibility of incorporating
the ideas raised into practical approaches to benefit-cost analysis.
OCR for page 14
14
INTRODUCTION
REFERENCES
Kneese, Allen V.
1989 The economics of natural resources. Pp. 281-309 in Michael S. I3eitelbaum and
Jay M. Winter, eds., Population and Resources in Westem Intellectual Traditions.
New York: The Population Council. Published as a supplement to Population
and Development Review, vol. 14. (Also reprinted as Reprint 243 in the RFF
Reprint Series, Resources for the Future, Washington, D.C.)
National Research Council
1983 Risk Assessment in He Federal Govemment: Managing the Process. Committee on
the Institutional Means for Assessment of Risks to Public Health. Washington,
D.C.: National Academy Press.
Quade, E.S.
1982 Analysis for Public Decisions. New York: Elsevier Publishers
U.S. Environmental Protection Agency
1987 E.PM. Use of Benefit Cost Analysis: 1981-1986. Office of Policy, Planning
and Evaluation, EPA-230-05-87-028. Washington, D.C,: U.S. Environmental
Protection Agency.
Representative terms from entire chapter:
regulatory decisions
