United States
Environmental Protection
Agency
Environmental Research
Laboratory
Corvallis, Oregon 97330
EPA-600/5-78021
October 1978
Research and Development
&EPA
SOCIAL DECISION-MAKING
FOR HIGH CONSEQUENCE,
LOW PROBABILITY
OCCURRENCES
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development. U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the SOCIOECONOMIC ENVIRONMENTAL
STUDIES series. This series includes research on environmental management,
economic analysis, ecological impacts, comprehensive planning and fore-
casting, and analysis methodologies. Included are tools for determining varying
impacts of alternative policies; analyses of environmental planning techniques
at the regional, state, and local levels; and approaches to measuring environ-
mental quality perceptions, as well as analysis of ecological and economic im-
pacts of environmental protection measures. Such topics as urban form, industrial
mix. growth policies, control, and organizational structure are discussed in terms
of optimal environmental performance. These interdisciplinary studies and sys-
tems analyses are presented in forms varying from quantitative relational analyses
to management and policy-oriented reports.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/S-78-d!21
October 1978
SOCIAL DECISION-MAKING FOR HIGH CONSEQUENCE,
LOW PROBABILITY OCCURRENCES
by
National Planning Association
Washington, D.C.
and
Haldi Associates, Inc.
New York, N.Y.
Contract No. 68-01-3228
Project Officer
John Jaksch
Criteria and Assessment Branch
Corvallis Environmental Research Laboratory
Corvallis, Oregon 97330
CORVALLIS ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CORVALLIS, OREGON 97330
-------
DISCLAIMER
This report has been reviewed by the Corvallis Environmental Research
Laboratory, U.S. Environmental Protection Agency, and approved for publica-
tion. Approval does not signify that the contents necessarily reflect\ the
views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
ii
-------
FOREWORD
Effective regulatory and enforcement actions by the Environmental Protection
Agency would be virtually impossible without sound scientific data on pollu-
tants and their impact on environmental stability and human health. Respon-
sibility for building this data base has been assigned to EPA's Office of
Research and Development and its 15 major field installations, one of which
is the Corvallis Environmental Research Laboratory (CERL).
The primary mission of the Corvallis Laboratory is research on the effects
of environmental pollutants on terrestrial, freshwater, and marine eco-
systems; the behavior, effects and control of pollutants in lake systems;
and the development of predictive models on the movement of pollutants in
the biosphere.
This study was initiated by the Washington Environmental Research Center,
Office of Research and Development, Washington, D.C. and completed at the
CERL, Office of Research and Development, Corvallis, Oregon.
A. F. Bartsch
Director, CERL
111
-------
PREFACE
Industrialized societies have long viewed technological innovation as a
primary means of obtaining ever higher levels of social well-being from
limited resources. Increasingly, though, new technologies bring with them
possible environmental consequences which are highly adverse, albeit with a
very low probability of occurrence: such technologies thus involve environ-
mental risk, a social problem that differs in nature from the familiar air or
water pollution problem, and is of growing importance.
Two technologies involving environmental risk have been selected for in-
tensive case studies in this report. They are: (1) widespread use of pesti-
cides which may have toxic, carcinogenic, mutagenic, or teratogenic effects;
and (2) disposal of nuclear wastes. Other examples of environmental risk
problems include: design and siting of nuclear reactors; use of nuclear ex-
plosive devices for stimulating the extraction of underground natural gas;
ozone depletion resulting from fluorocarbon emissions; exploitation of subsur-
face oil, especially near scenic coastlines,such as Santa Barbara; adoption of
drugs having a small probability of extreme biological damage, such as thalid-
omide; dangers that experiments with recombinant DNA will create a new disease
against which mankind has no natural defense; and the shipment and handling of
highly flammable liquified natural gas.
A number of characteristics distinguish problems o£ environmental risk
from environmental pollution problems: (1) The probability of occurrence is
so low and the underlying mechanism often so uncertain that even the most in-
formed estimates of risk are highly subjective and vary widely. (2) The
level of potential damages is frequently catastrophic and overwhelms the
underwriting capability of the private insurance mechanism, or at times even
the possibility of reasonable public provision. (3) Many environmental risks
are latent; they may be borne unknowingly and involuntarily for long periods
of time. (4) The consequences of some environmental risks, once incurred,
cannot be reversed for long periods of time—in the case of nuclear wastes the
risk may extend for tens of thousands of years. (5) Environmental risk situ-
ations typically involve collective exposure of large segments of society.
(6) Parties who bear environmental risk typically receive a disproportion-
ately small share of the benefits from the undertaking. Finally, (7) for many
risks it is difficult to establish a legally defensible cause-and-effeet rela-
tionship between the parties liable for the event and those who bear the
losses.
The preceding characteristics make environmental risk problems hard to
manage through existing regulatory, legal and economic institutions. Technol-
ogies that involve environmental risk thus face society with extremely diffl-
iv
-------
cult collective decisions. Both the liability issue and the magnitude of po-
tential adverse consequences raise a number of important social questions
such as: Does the activity infringe civil liberties or other constitutional
guarantees? To what extent is the public willing to trust the risk assess-
ments of those who are promoting the activity and are most likely to benefit
from it? What should decision makers do when recognized authorities give
widely varying estimates of risk and are in almost total disagreement con-
cerning recommended action? How can we arrive at a socially acceptable de-
cision? Are there "optimal" decisions? Will decisions that are optimal by
an "objective" cost-benefit criterion also be socially acceptable? What is
the risk of adopting decisions or policies that are not socially acceptable?
Is sabotage, revolution and possible destruction of democratic society at
risk? If social acceptability is an important criterion, how does one im-
prove the decision process so as to make decisions more acceptable? It is
questions such as these which have motivated this study, and which it ad-
dresses to the extent possible.
Events of the nineteen seventies have illustrated again and again the
need for a better understanding of these issues. Example: controversy con-
cerning the construction of nuclear plants has not diminished, despite in-
creased experience with nuclear power. Even in regions where energy is in
short supply and the cost is high, such as New Hampshire, thousands of people
regularly turn out to protest a nuclear plant whose construction is well-
advanced. Another example: a large number of scientists have proposed estab-
lishment of a new institution, to be called a "science court," because they do
not have sufficient confidence in the existing decision processes. Another
example: in Japan thousands of demonstrators get involved in sabotage, mass
demonstrations and mass arrests to prevent opening of the new Tokyo airport,
because the decision process is not socially acceptable.
Economists, decision theorists and political scientists have made con-
siderable progress in applying analytical techniques to the process of social
choice. However, these techniques rarely focus on situations involving low
probability, high consequence outcomes. Decisions involving risk and uncer-
tainty have been studied, but principally in the context of international
policy and military situations. More work is needed on how to arrive at so-
cially acceptable choices when a technology involves low probability but very
high consequence outcomes.
The study team was aided greatly throughout the project by an Advisory
Panel composed of Kenneth Arrow, Otto Davis and Roland McKean. To these three
individuals we wish to express our sincere appreciation for their helpful ad-
vice and comments. The authors also wish to thank the many individuals from
the Environmental Protection Agency who cooperated with us during the study.
The study team for this project consisted of: Mark Kendall and Ivars
Gutmanis from the National Planning Association; and John Haldi and Thomas
Vietorisz from Haldi Associates, Inc. Primary responsibility for individual
components of this study was as follows. Principal investigator for Part I,
Theoretical Aspects: Mark Kendall (with the exception of the subsection in
Chapter 3 entitled "The Pareto Criterion," which was written by Kenneth
Arrow). Principal investigator for Part II, Case Study of Pesticides: John
v
-------
Haldi. Principal investigator for Part III, Case Study of Nuclear Wastes:
Ivars Gutmanis.
To all those who offered assistance, advice, and helpful criticism, the
authors wish to express their thanks. Any errors or shortcomings of the study
are, of course, their responsibility alone.
The data, analyses and conclusions in this report are solely the re-
sponsibility of the authors and are not to be interpreted as expressing those
of the National Planning Association or of the members of its Board of
Trustees, committees and staff.
vi
-------
ABSTRACT
This study deals with the process of reaching social decisions that in-
volve low-probability, high-consequence outcomes. It is divided into three
major parts:
I. Theoretical considerations
II. Case study of chlordane and heptachlor pesticide
regulation
III. Case study of nuclear waste disposal
Part I reviews the two main classes of criteria which have been proposed
for such social decisions: (1) approaches based on the market mechanism and
its extension, cost-benefit analysis, which is the operational method of im-
plementing market-based criteria where full market information is lacking; and
(2) approaches associated with Rawls and Buchanan, which focus not only on
outcomes, but also impose a set of minimal constraints on the process whereby
decisions and social consensus are reached. While formal optimality (in a
Pareto or cost-benefit sense) is important, social acceptability of the pro-
cess used to arrive at decisions is seen to be equally important. Part I con-
cludes by proposing a set of eight criteria for evaluating a social decision
process. These criteria are capable of being applied to a wide range of deci-
sion processes, be they deliberations before Congress, the Courts, an adminis-
trative law judge, the National Academy of Sciences, or an internal executive
process. Applicability of the criteria is then explored in the context of
two case studies.
The pesticide case study focuses on the chlordane/heptachlor suspension
hearings of the U.S. Environmental Protection Administration as an example of
the functioning of one particular decision process, the administrative law
procedure. Applicable law mandates that the Administrator base his decision
concerning use of a given pesticide on a cost-risk-benefit criterion. Since
such analysis is not, per se, fully understood or accepted by many segments of
society, the law also prescribes that the analysis be subjected to an adver-
sary procedure conducted before an administrative law judge. The legal pro-
cess is thus used as a forward-looking decision process in an effort to in-
crease social acceptability of the outcome.
The case study on nuclear wastes examines a distinctly different decision
situation. While EPA has sole responsibility for pesticide decisions, no
agency has sole responsibility for the disposal of nuclear wastes. Certain
key decisions in this area still remain with Congress. The Nuclear Regulatory
Agency and Department of Energy have inherited major roles in the regulation
of nuclear power, but many other agencies are also involved one way or another
in decisions concerning or affecting nuclear wastes. This decision process is
poorly articulated. The eight criteria proposed in Part I nevertheless pro-
vide a useful tool for evaluating the process.
vii
-------
This report was submitted in fulfillment of Contract #68-01-3228 by the
National Planning Association (acting as prime contractor) and Haldi Asso-
ciates, Inc. (acting as subcontractor), under the sponsorship of the U.S.
Environmental Protection Agency. This report covers a period from 6/19/75 to
4/30/77, and work was completed as of 6/1/77.
viii
-------
CONTENTS
Foreword
Preface iv
Abstract vii
Figures x
Tables xi
PART I: THEORETICAL CONSIDERATIONS
1. Introduction 1
2. Distinguishing Features of High Consequence,
Low Probability Occurrences 3
3. Alternative Approaches to Social Decision-
Making 11
4. Criteria for Evaluating Social Decision
Processes 29
References for Part I 37
PART II: CASE STUDY OF CHLORDANE/HEPTACHLOR
5. Introduction 41
6. Review of Federal Pesticide Regulation 45
7. Issues Raised in the Chlordane/Heptachlor
Suspension Case 59
8. Theoretical Issues Posed by Potentially
Carcinogenic Pesticides 73
9. Broader Issues Concerning the Pesticide
Decision Process 99
10. Evaluation of the Pesticide Decision Process Ill
References for Part II 125
PART III: CASE STUDY OF NUCLEAR WASTE DISPOSAL
11. Introduction 133
12. Disposal Concepts for Nuclear Wastes 141
13. Risk Issues Associated with Nuclear Waste
Management 159
14. Legal Framework and Federal Criteria for
Nuclear Waste 165
15. Publicly Voiced Criteria for Nuclear Waste
Material 179
16. Evaluation of the Decision Process as Applied
to Nuclear Waste Management 189
References for Part III 201
ix
-------
FIGURES
Number Page
1 Interrelationship Between Scientific Functional
Categories of a Hazard Evaluation System 84
2 Generation of Nuclear Waste Within Nuclear Fuel Cycle:
Major Options for Waste Management 142
3 Geological Disposal from Geosphere Migration
Point of View 161
4, Comparison of Characteristic Natural Ages with Control
Time of High-Level Waste 163
5 Hazard of Spent Fuel as a Function of Age 164
-------
TABLES
Number Page
1 Proposed Evaluation Criteria and Their Derivation 32
2 Uses of Chlordane and Heptachlor, 1973-1974 60
3 Oxychlordane Residue Found in Human Adipose Tissue, 1971-1974. ... 62
4 Heptachlor Epoxide Residue Found in Human Adipose
Tissue, 1970-1974 62
5 Criteria for Evaluating Decision Processes Embodying Low
Probability, High Consequence Risks 113
6 Typical Materials in High-Level Liquid Waste 144
7 Characteristics of Solidified High-Level Waste 146
8 Summary of Nuclear Waste Disposal Concepts 148
9 Characteristics of Geologic. Disposal Concepts 151
10 Extract of Results of Open-ended Questioning on "The Two or
Three Main Disadvantages of Nuclear Power Plants to Produce
Electric Power" 183
11 Agreement with Stated Problems Connected with Nuclear Power
Plants 184
12 Criteria for Evaluating Decision Processes Embodying Low
Probability, High Consequence Risks 195
xi
-------
Part I
THEORETICAL CONSIDERATIONS
-------
1. Introduction
A category of environmental problems of increasing importance involves
situations characterized by high consequence outcomes which have a low proba-
bility of occurring. Examples of situations involving such environmental
risk include; (1) the risk of ozone depletion due to emissions of fluoro-
carbons; (2) the risk of radioactive leakage from nuclear accidents or from
disposal of nuclear wastes; (3) the risk of creating a new disease, for which
there is no known cure or vaccination, from experiments with recombinant DNA;
or (4) the risk that residues for various pesticides or other chemicals may
be carcinogenic, mutagenic or teratogenic.
Environmental risk problems involve a number of considerations that
differ substantially from what might be described as classical or traditional
environmental problems of air and water pollution. For example, uncertainty
regarding events and outcomes is somewhat less in classical pollution prob-
lems. That is, if certain chemical wastes are dumped into rivers, the down-
stream effects on fish or oxygen content can now be predicted (and measured)
with a fair degree of certainty. The uncertainty attending situations
described above as environmental risk is greater by at least an order of
magnitude. This uncertainty, coupled with several other important distin-
guishing characteristics, makes problems of environmental risk far less sus-
ceptible to management through existing social decision processes. Debate
and disillusionment concerning our decision process already exist. It is
important, therefore, that we develop explicit criteria for evaluating the
decision process itself.
In the context of this study, the social decision process is defined
broadly. It includes the legislative process in Congress and the various
regulatory processes established by Congress, (e.g., EPA, FDA, etc.),
including judicial review of administrative decisions. Also included, to
some extent, are the deliberations and procedures of the National Academy of
Science. While not an official part of the decision process, the National
Academy is a well-established institution whose opinions weigh so heavily
with elected officials that it is accorded at least semi-official status.
Moreover, some of the Academy's procedures have been criticized by a number
of scientists, some of whom have recommended that the Academy's deliberations
be replaced, at least in part, by a new institution referred to in the liter-
ature as a Science Court.
By way of overview, Part I of this report addresses some fundamental
considerations that confront the decision process. The purpose is to develop
a set of criteria for evaluating alternative decision processes. In Part II,
a case study of chlordane and heptachlor includes a detailed description of
the social process for regulating pesticides. The principal concern with low
-------
probability, high consequence events is that use of some pesticides may cause
unintended, widespread effects, such as cancer or animal kills. Pesticide
regulation is primarily the responsibility of the Environmental Protection
Agency. Nuclear waste disposal, in Part III, is the other case study. Ob-
viously, waste disposal is an integral part of the nuclear fuel cycle. Use
of nuclear power and accompanying problems of waste disposal have been sub-
ject to wide public debate. The agency primarily charged with making social
decisions concerning the appropriate method of waste disposal is the Nuclear
Regulatory Commission (a successor to the Atomic Energy Commission). In
contrast to pesticides, however, the current decision process for nuclear
wastes is far more diffuse.
The next chapter summarizes the important characteristics of social
decisions involving low probability, high consequence outcomes as exhibited
in the two case studies. Chapter 3 discusses various approaches to evalua-
ting social decisions with the unique characteristics described in Chapter 2.
Our criteria for evaluating social decision-making processes involving low
probability, high consequence occurrences are developed in Chapter 4. These
criteria are subsequently applied in Chapter 10 to the pesticide regulatory
process; in Chapter 16 they are applied to nuclear waste disposal decisions.
-------
2. Distinguishing Features of High Consequence,
Low Probability Occurrences
Social decision-making involving low probability, high consequence
occurrences has a number of interesting and difficult aspects. These are
discussed in detail in the two case studies which accompany and are an inte-
gral part of this report. Succinctly, special features of environmental risk
problems developed in the case studies are:
1. Aggregation of preferences
2. Limited knowledge
3. Uncertainty and irreversibility
4. Intergenerational effects
5. Distribution of benefits vjs^ risks
6. Counter risks
Each of these features raises theoretical issues that need to be resolved by
an acceptable social decision process.
Aggregation of Preferences
Most individuals are risk adverse. This implies that the satisfaction a
person receives from a certain income of $10 exceeds the satisfaction of a 50
percent chance of a zero income and a 50 percent chance of a $20 income. i_/
Thus, it is necessary to know preference functions (i.e. , risk valuations) as
well as possible outcomes when evaluating social decisions which entail low
probability but high consequence outcomes.
Limited Knowledge
Decisions concerning pesticides and nuclear waste disposal test the
limits of our knowledge. We know for instance that chlordane and heptachlor
persist in the food chain. Evidence also indicates that they are carcino-
genic in rats, but we do not know the dose-response pattern with any signifi-
cant degree of certainty. The suspension hearing for chlordane and hepta-
chlor established that there is reason to believe that continued use of these
two pesticides may result in a higher incidence of cancer among humans. How-
ever, it was not possible to establish the relationship between specific use
levels of chlordane and heptachlor and specific increases in the incidence of
either cancer in general, or among specific groups who may be more suscep-
tible to the disease. Also, it was not possible to specify with certainty
the increase in use of substitutes for chlordane and heptachlor and the sub-
stitutes' carcinogenic effects. This uncertainty — lack of knowledge — is
particularly important because chlordane and heptachlor are effective
-------
pesticides. Lack of knowledge makes difficult the application of an effi-
ciency criterion (or any other standard).
Knowledge gaps for nuclear waste disposal are probably larger than for
pesticides. There does not exist an accepted technology and set of social
institutions that can assure a well-defined distribution of risks associated
with disposing of nuclear wastes. Some are willing to assume, given man's
historical inventiveness, that suitable technologies and institutions will be
developed; they urge continued development of nuclear power. Others are un-
willing to make this assumption. They argue that risks associated with all
present proposals for waste disposal are either too high or are unknown.
This is not dissimilar to the situation encountered with pesticides. Bene-
fits of increased nuclear power, like the benefits of chlordane and hepta-
chlor for increased food production, can be reasonably well-specified, but
the link between increased use of nuclear power and nuclear "pollution", like
the link between continued use of chlordane and heptachlor and cancer, is not
well-specified. Even in a world where the social decision maker knew every
individual's risk preference, the limits of knowledge would prevent a suit-
able aggregation of these risk preferences, since we do not know the distribu-
tion of risk.
Finally, infrequent observation of a bad occurrence, implied by low
probability, makes it difficult to determine the correct response to a bad
occurrence. A large number of experiences with a bad occurrence generally
provide sufficient information to learn the "best" response. For low proba-
bility events, circumstances of a bad occurrence are unknown, and hence,
possible responses are unknown. Finally, a low probability, high consequence
occurrence makes any type of learning (observation or experimentation) very
expensive.
Uncertainty and Irreversibility
In most cases, uncertainty comes from two sources. First is lack of
knowledge. For example, there are significant differences of opinion con-
cerning the long-term implications of various forms of nuclear waste dis-
posal. Second is the pure probabilistic nature of many problems requiring
social decisions that involve low probability, high consequence occurrences.
In a world of perfect knowledge, certain components of the nuclear waste dis-
posal problem would remain probabilistic since nature is, in part, stochastic.
Assume for the moment that decisions are evaluated as to whether they
approximate the competitive market outcome. With given expectations, at any
point in time the competitive market tends to balance the otherwise conflic-
ting expectations of agents in the economy. One individual's expectations
concerning future prices and states of the world might be significantly dif-
ferent from another's. Also, attitudes toward risk may differ significantly.
For example, expectations of all agents in an economy are not the same nor are
their risk preferences identical. However, the private economy has histori-
cally been able to find a set of prices and quantities that balance conflic-
ting expectations and attitudes toward risk.
-------
If an individual sees a set of prices and quantities that he did not
expect, he revises his expectations, and a new equilibrium may be reached.
Agents adjust their holdings of stocks, bonds, commodities and other goods
in response to their revised expectations. They may "hedge their bets."
They may suffer windfall monetary losses or gains, but seldom do they experi-
ence a windfall loss or gain that is not reversible at some cost. In the
absence of externalities, the impact is on the agent making the decision, not
on others.
One of the major differences between this private balancing of expecta-
tions and a social decision concerning nuclear waste disposal or the use of
chlordane and heptachlor, is that many of the possible outcomes are not re-
versible. If storage of nuclear waste in salt mines does, through some
unknown means, pollute the Western watershed, we can revise our expectations
concerning the feasibility of nuclear power but we cannot reverse the impact
of our incorrect expectations. 2J Likewise, the discovery that the use of )C
pounds of chlordane per year increases the incidence of cancer by Y_ percent
will lead to a revision in our expectations. However, given the pervasiveness
of pesticide residues in the food chain, the impact of our initial decision
may be irreversible for many people. Thus, the consequence of a bad occur-
rence may not be technically reversible, or may only be reversible over such
a long period that, given some positive rate of time preference, it may as
well not be reversible.
The private market economy, with the help of bankruptcy law protection,
permits individual economic agents to revise expectations and, in at least
some sense, reverse their losses, for a price within a reasonable time period,
if they are wrong. This reversibility permits the efficiency criterion to
maintain its validity as a suitable framework for analyzing competitive mar-
kets. The irreversibility of certain decisions concerning low probability,
high consequence occurrences coupled with sensitivity to expectation changes
(see Chapter 3) raise significant questions concerning validity of the
efficiency criterion for evaluating these social decisions.
Intergenerational Effects
The requirement to consider intergenerational effects is important given
the effect of uncertainty and 80,000 year half-lives. The benefits of nuclear
energy are significant. It may be relatively cheap, it relieves our depen-
dence on foreign sources, and it would be an almost inexhaustible source of
energy with development of the breeder reactor. However, waste disposal,
given present and reasonably foreseeable technologies, requires "a diligence
and longevity of our social institutions that we are quite unaccustomed
to." 3/
Consideration of intergenerational effects implies that the most sensi-
tive variable for any social decision, using traditional benefit-cost analy-
sis, is the interest rate. The literature is replete with discussions about
the "optimal" social interest rate. These arguments usually search for an
interest rate that will assure that decisions concerning investments in public
capital will satisfy the Pareto criterion by being consistent with the private
and public markets. Taxes and other considerations generally imply that an
-------
appropriate social interest rate is lower than the corporate return on
investment. 4_/ An obvious question is whether these benefit-cost-risk
assessments are accurate. Even if they are, the uncertainty of nuclear
waste disposal procedures and the associated intergenerational comparisons,
may imply that these assessments, with their implied acceptance of an effi-
ciency criterion, are not sufficient. For example, Kneese states:
It is my belief that benefit-cost analysis cannot
answer the most important policy questions associated
with the desirability of developing a large-scale
fission-based economy. To expect it to do so is to
ask it to bear a burden it cannot sustain. . . These
questions are of deep ethical character. 5f
The ethical character of the nuclear debate is whether (or, under what
conditions) the present generation has the right to impose on future genera-
tions the uncertainties associated with a fission-based economy. If we do
not know all the risks involved and if we do not know the preferences of
future generations, do we have the right to impose these uncertainties? The
responsibility for caring for the wastes generated by a fission-based economy
will fall to future generations. Does the present generation have the right
to impose this responsibility? If it does, under what conditions?
If one accepts the concept of maximizing expected utility over time, with
time partitioned into today and tomorrow, then the real interest rate observed
today is composed of:
(a) pure time preference;
(b) a risk premium; and
(c) after removing time preference and the risk premium,
a factor that reflects the real rate of economic growth.
In order to develop a suitable social interest rate, it is necessary to
examine each of these components. It has long been argued, on grounds of
rationality as well as ethics, that there should be no pure time preference
in the absence of uncertainty concerning death. Sidgwick maintains that, in
the absence of uncertainty, rationality implies an impartial concern for all
parts of our life. Mere differences in time would be irrelevant when evalua-
ting the utility of goods at various points in time. If the universal good
and individual good are similar in essential respects, social time preference
should be zero. 6_/
Since the risk premium is most difficult for our purposes, we will con-
sider the third element next. If future generations are expected to be
"better off" due to technological change and capital accumulation, then it is
rational and ethical to discount future values at a rate that equates the
utility offered by the total output of the economy at each point in time. If
the goods produced next year will give two percent more total utility than the
goods produced this year, then fairness requires that sacrifices in utility
today be balanced against future utility increases. Thus, economic growth
implies that the third component should be positive.
-------
The second component in the observed market interest rate, the risk pre-
mium, is particularly difficult to handle in social decisions for low proba-
bility, high consequence occurrences. For some social investment decisions,
the risk premium can be ignored since the benefits and costs of any one social
investment are small relative to the wealth of any one individual, which off-
sets any substantial risk aversion held by members of society. 7j However,
risks arising from nuclear waste disposal or pesticide residues can possibly
be large to all members of society, present and future. We do not know the
shape of present or future utility functions, and hence attitudes toward risk,
although it is necessary to estimate an appropriate risk premium in order to
employ the usual benefit-cost-risk analysis. This requirement is probably so
substantial as to make it impossible to use, unhesitatingly, the usual
benefit-cost-risk analysis in making social decisions for low probability,
high consequence occurrences. The result of any long-term benefit-cost
analysis is sensitive to the interest rate, and reasonable bounds on a risk
premium may not exist, since the risks are not independent over time.
Distribution of Benefits vs. Risks (costs)
A benefit-cost analysis of a social investment usually attaches no expli-
cit value to the distributional consequences of the investment. It is often
argued, however, that these effects should be taken into account "on the
side." For both pesticides and nuclear wastes, the re-distribution of costs
and benefits may be substantial and asymmetric. Also, restricting pesticides
or preventing waste disposal hazards might yield benefits that differ geo-
graphically as well as temporally. In the case of pesticides, the benefit of
limiting a pesticide with persistent residues and probable carcinogenic
effects is generally assumed to accrue to all of society. At first blush,
this assumption may seem appropriate. However, if one admits differential
values of human life and assumes that these differentials are positively re-
lated to future, expected income streams, then the value of regulating a
pesticide will accrue to those with higher expected income streams. (Or, if
people value risks differently, regulation of a pesticide will give greater
benefits to those who are more risk adverse.)
The cost of regulating or discontinuing the use of a pesticide generally
accrues to (1) landowners or owners of specialized farm resources, to the ex-
tent that they cannot shift their increased costs of food production to the
consumer, and (2) to the consumer, to the extent that the increased costs of
production can be passed along. If, for simplicity, all increased costs were
shifted completely to the final consumer, all consumers would face roughly
the same increases in food cost. However, in terms of welfare loss, those
consumers with relatively lower incomes, and hence a relatively higher portion
of their budgets devoted to food, would experience a larger welfare loss. 8_/
Thus, in this admittedly simple model, the primary beneficiaries of limita-
tions on the use of pesticides are the higher income groups.
A similar asymmetric distribution of costs and benefits can be attributed
to limitations on nuclear waste disposal, since those who have higher expected
income streams would tend to offer more for decreased threats to their income
stream from unforeseen consequences of nuclear waste disposal. 9/ Also, a
-------
higher price of energy, because energy has a relatively low income elasticity,
would adversely affect lower income groups.
Inequities of the intragenerational distribution of benefits and costs,
for both pesticides and nuclear waste disposal, might be corrected by changes
in the tax system. However, if the continued use of pesticides or the in-
creased use of nuclear power does offer a significant threat to society as we
presently know it, then future groups might suffer a higher cost for deregula-
tion of pesticides and increased nuclear waste disposal, relative to the pre-
sent generation. This could not be easily corrected by changes in the tax
system. Thus, there is need for a more explicit consideration of the intra-
generational and intergenerational distribution of benefits, costs and risks.
Counter Risks
For both case studies, one side of the benefit-cost relationship seems
reasonably well-defined. In the case of pesticides, a primary risk consider-
ation is persistence of the pesticide residues and an associated increase in
the incidence of cancer. Benefits of decreased use of chlordane and hepta-
chlor -are uncertain, while it is generally agreed that the costs of decreased
use of these pesticides are fairly well specified. It is generally accepted
that, through the use of substitute chemicals, the costs of decreasing the use
of chlordane and heptachlor are small relative to expected benefits (except
for the possible carcinogenic effects of substitutes).
If regulation and limitations on the use of pesticides continue, it is
reasonable to speculate on longer term uncertainties associated with the cost
estimates. For example, pesticide regulation is sequential. That is, each
pesticide is considered in turn, and the costs of its regulation are estimated
assuming no further regulation of pesticides. If, over time, increased limi-
tations on other pesticides effectively limit the ability to control particu-
lar pests, the costs of limiting the use of chlordane and heptachlor might be
much larger, after other pesticides are limited in their use, than when the
initial decision was made. More serious, substitutes for chlordane and hepta-
chlor might have more detrimental effects than either of the two original sub-
stances. Thus, if chlordane and heptachlor are not available to substitute
for the possibly more dangerous pesticides, the present regulatory mechanism
may overestimate the costs of regulating the latter. In other words, each
sequential decision might be valid, while the total set of decisions is not,
since pesticides prohibited earlier in the sequence are not considered
"available" substitutes.
The problem of risks "on the other side of the equation" is more sub-
stantial for nuclear power. Whereas costs and associated risks of waste dis-
posal are subject to a high degree of uncertainty, many individuals, at least
until recently, have been willing to concede the accuracy of order-of-
magnitude estimates of the benefits from nuclear power. Since establishment
of nuclear power requires a relatively long lead time and generates relatively
long-lived plants, most benefits of nuclear power will accrue at some future
date. Obviously, estimating benefits requires a fairly precise estimate of
the future demand for energy. If this estimate cannot be made with a reason-
able degree of precision, risk-assessment problems occur in estimating the
-------
benefits of nuclear power. This estimate also requires assumptions concern-
ing both the future state of world (oil) diplomacy and the availability of
substitute energy sources.
Summary
This chapter has focused attention on several important characteristics
that distinguish problems of environmental risk from the more traditional
pollution problems. Classical benefit-cost-risk analysis can guide social
decisions for low probability, high consequence occurrences. However, the
decision process must also consider (1) factors that, due to measurement
problems, cannot be included in the quantitative benefit-cost-risk analysis;
and (2) ethical elements such as preventing the tyranny of one group over
another (or, one generation over another).
Notes to Chapter 2
1. Milton Friedman and L. J. Savage, "Utility Analysis of Choices Involving
Risks," Journal of Political Economy. Vol. 56, (August 1948), pp. 279-
304.
2. This also limits the use of a Baysian approach for estimating the proba-
bilities since (1) bad occurrences happen infrequently, which gives few
observations for estimating and revising probability estimates, and (2)
irreversibility implies that a bad occurrence limits the use of a new
Baysian probability estimate.
3. Alan M. Weinberg, "Social Institutions and Nuclear Energy," Science
(July 7, 1972), p. 31.
4. For example, see Arnold Harberger, "Professor Arrow and the Discount
Rate" in Project Evaluation. (Chicago: University of Chicago Press,
1976).
5. Alan B. Kneese, "The Faustian Bargain," Resources (September 1973), p. 1.
6. See John Rawls, A Theory of Justice, pp. 293-94 for this interpretation
of Sidgwick.
7. See Kenneth J. Arrow and Robert C. Lind, "Uncertainty and the Evaluation
of Public Investment Decisions," American Economic Review (June 1970),
pp. 364-78.
8. If all income groups have the same price elasticity of demand for food.
9. If bad occurrences from nuclear waste disposal are limited geographically,
the greater mobility of higher income groups would decrease the value they
place on decreased generation of nuclear waste.
-------
3. Alternative Approaches to Social Decision-Making
A fair spectrum of criteria for evaluating decisions of social institu-
tions ranges from (1) Pareto optimality, constrained by an appropriate initial
distribution of endowments, to (2) Rawls' maximin principle, constrained by
equality of opportunity for all, to (3) Buchanan's claim that social institu-
tions are just if they are entered into freely by individuals. These three
sets of criteria do not describe the complete spectrum. _!/ They are suffici-
ently diverse, however, to provide a good basis for developing criteria that
can be used to evaluate the social decision process. The following discussion
summarizes these procedures.
The Pareto Criterion
Pareto optimality or efficiency requires that all agents of the economy
be brought into trading arrangements and that trades take place until there
are no more mutual gains from further trade between the agents. The Pareto
criterion provides the rationale for a market-based decision process. For
nonmarket decisions by government, the Pareto criterion is the rationale
underlying the widespread use of benefit-cost analysis. The trading economy
is said to be efficient (or optimal) if the final exchange of goods is a point
on the familiar contract curve. This result holds with the usual assumptions
of continuity, insatiability and transitivity of preference orderings.
Neoclassical analysis can be used to define Pareto efficient decisions
under a variety of conditions. However, the unique characteristics of envi-
ronmental risk problems are only beginning to be widely recognized, and Pareto
efficient decisions have not been fully worked out for such situations. The
following notes represent tentative steps towards a theory of evaluating
publicly manageable risks to life. They are considered to be correct insofar
as they go, although they clearly leave out a large number of problems that
arise in practice.
The static case: objective function and instruments. The individual is
supposed to have a risk of dying. If he survives, he consumes and derives
satisfaction from his consumption. According to the usual expected-utility
theory, his choices are governed by the desire to maximize,
(1-p.) U. (consumption) + P± U(death), (1)
where,
P. = probability of death, U. = utility of individual i.
11
-------
As a first approximation, it seems reasonable to consider the utility of
death to be a constant, which, without loss of generality, we may take to be
zero. There are two possible reservations to this assumption: (1) the
manner of death may matter to the individual, so that he may be willing to
devote his own resources to changing the conditions under which death occurs
or to protecting against one form of health hazard even if it only makes in-
evitable another form; (2) care for the family or others means that, in event
of death, utility depends upon the bequest or insurance protection. Both
points can be met by extension of the present arguments; but, in any case, these
are both probably second-order considerations. Setting the utility of death
equal to zero means that utility functions are not only cardinal, as in the
usual expected-utility theory, but have also a fixed origin, hence unique up
to positive multiplication (what the psychologists call ratio scales). Then
the maximand is,
(1 - Pi)ui(ci)» (2)
where
c. = consumption of individual i. (3)
The novel aspect of resource allocation in the theory of optimal safety
is that the probability, p^, of death is itself controllable by allocating re-
sources to it. The general assumption to be made here is that the probability
can be affected by both public and private measures. This is clearly realis-
tic: the probability, pi} is the resultant of a great many separate process-
es, some essentially private in nature, some public (in Samuelson's sense).
Let,
x = public expenditures on decreasing probability of death,
s. = private expenditure of individual i on decreasing probability
of death, (4)
and we assume,
Our aim is to define a Pareto optimum of the maximands (2) with the
assumption (5). We shall assume that resources are freely transferable be-
tween the public and private sectors and that, in any case, each individual
will choose optimally in allocating his income between c. and s..
To illustrate the analysis in the simplest case, assume first that there
are no private effects on the probability of death, i.e., p. is a function of
x alone and not of s.. Our resource constraint then is,
£c. + x given. (6)
12
-------
In the usual way, we take arbitrary positive weights, w., for the maximands
(2) and optimize for each such choice. The Pareto frontier is found by
considering the allocations for all possible w *s; equivalently , we eliminate
the w.'s from the optimum conditions.
Maximize,
Iw± [1 - PjL(x)] VG..),
subject to (6). Let y be the Lagrange parameter corresponding to (6). Then
differentiation of the Lagrangian with respect to x yields,
-EWi(dPi/dx) ^(c..) = y, (7)
while differentiation with respect to c. yields,
w±[l - pi]U^(ci) = u
Multiply the i term on the left-hand side of (7) by,
1 = y/Wj.OL - P±) Ui,
which clearly leaves it unchanged in value; then divide the resulting
expression derived from (7) by
dpi U
- , _ =
i 1 - p. dx ' U.
*i i
t
To interpret (9), it is useful to interpret the factor U /U . Start
with a consumption level c. and probability of death, p.. If consumption is
increased by an amount h, the individual will remain on the same indifference
curve with a slightly higher probability of death; let that probability be
p.(h), where, p.(0) = p.. Then
[1 - P±(h)] l^Cc.. + h) = (1 - PI) ILCc^.
Differentiate with respect to h:
- p^(h) U..^ + h) + [1 - PjL(h)] V±(c± + h) = 0
Set h = 0.
(1 - P±) IL(CI) / U±(c±) = p^(0). (10)
13
-------
Note that p (0) is the marginal increase in probability of death per unit in-
crease in consumption. Its reciprocal would be the marginal compensation
needed to offset a unit increase in probability of death, or, alternatively,
the marginal expenditure the individual would be willing to make to reduce
the probability of death (per unit reduction). Let, then,
CJ(PJ) = consumption needed to compensate for change in p., (11)
where the movements are along some specified indifference curve; then,
and (10) can be written,
(1 - P±) v[(
Substitution into (9) yields the perhaps unsurprising result ,
-JcJ.Cp±) (dp/dx) = 1; (12)
the public expenditure, x, should be carried up to the point where the
changes in individual probabilities of death, weighted by the individuals'
monetary values of the change, sum to 1.
i
So far, there has been no way of determining the valuations c. objec-
tively. Let us now suppose, as is reasonable, that in fact the probability
of death may be affected by private as well as public decisions. In addition
to consumption for each individual, there are also his expenditures on safety.
Let,
s, = safety expenditures of individual i, (13)
and we assume now,
pi = Pi^si'x^'
To find a Pareto optimum, we maximize,
subject to,
l(c± + si) + x given. (14)
14
-------
Let V again be the Lagrange parameter. Differentiation with respect to x,
ci, s.^, respectively yields,
-Zw±(3Pl/3x) U±(Ci) = (15)
w [1 - p] U(c) = u, (16)
-w± Op1/3s1) Ui(ci) = u. (17)
Note that equations (15) and (16) are identical with (7) and (8) except
that the total derivative, dp,/dx, has been replaced by a partial derivative,
9p./3x. Hence, the subsequent manipulations leading to (12) remain valid:
-£C^(PI) (3Pl/8x) = 1. (18)
On the other hand, from (16) and (17), we have, by eliminating and
dividing through by w . ,
9pi/9s1 = (1 - P±) U(ci)/Ui(ci). (19)
The derivative, 3p./3s. , represents the effect of private expenditures
on the probability of death. In the original definition, p. = p (s , , x) , it
may be useful to solve for s. in terms of p. and x; this function represents
the amount of private expenditures needed to achieve a given probability of
death, for given public expenditures. Then, 3s. /3p. is the marginal cost to
the individual of a unit change .in the probability of his death. By the
usual inverse function theorems,
3s±/dp± = l/OPl/3s±).
Substitute into (19) and also use the equation preceding (12).
(20)
at an optimum, the marginal willingness to pay for a unit reduction in prob
ability of death should equal the marginal cost, of such an .improvement in
terms of individual expenditures.
Notice also that an individual will, in allocating his own budget opti
mally, realize (20). Hence (20) provides a means of estimating from
objective data the marginal values of decreases in probability.
15
-------
The static case; several kinds of public safety expenditures^ Suppose
the probability of death for any individual can be affected in several ways
by public expenditures, thus, by increased plant safety standards paid for by
the government, by increased highway expenditure, or by increased police
force. We assume,
Pi = Pi(si' xl> X2) (21)
Then essentially the same results follow, with (15) being valid separately
for x1 and x2, and therefore (18) replaced by the two conditions,
-£c^(p±) Op±/8xj) = 1 (j = 1,2). (22)
Subtracting the equation for j = 1 from that for j = 2 yields,
-£ Cj(P±) [OPi/8x2) - (3pi/3x1)] = 0 (23)
The savings in probabilities of death attributed to each kind of expen-
diture, weighted by individuals' willingness to pay for safety should be
equal. This provides a basis for judging the level of a new kind of public
safety expenditure by reference to old ones.
There is a widespread argument for neglecting distributional effects
in benefit-cost analysis on the grounds that the effects of the particular
project under consideration are distributed in the population independently
of the marginal social value of a dollar. Analogously, one could assume that
the willingness to pay for safety is distributed among individuals inde-
pendently of the differential effect of one public expenditure as against
another. By elementary rules of statistics, independence implies that the
left-hand side of (23) can be written,
- (1/N) [Z c^(p±)] Z [Op1/3x2) - Opj/ax^],
and since certainly,
I c'(p )> 0,
i x 1
it follows from (23) that, if willingness to pay for safety is distributed
in the population independently of the safety effect of shifting expenditures
from one type to another, then optimality requires that,
E Op /3x,) = Z (3p./8x9),
± 1 J. ± 1 2.
i.e., so as to maximize the expected number of lives saved.
For its interest, we present the special case where each safety expendi-
ture corresponds to an independent risk of death. That is, the probability
16
-------
of survival is the product of the probabilities of survival from different
causes, each affected by a different category of expenditure.
1 0
1 - P. = [1 - p.(s.)J [1 - p.(x,)] [1 - p . (x ) ]. (24)
Then, by differentiation with respect to the variables s., x.. , x~, and
multiplication by -1, we see,
i o
OP,. - 1 - P • dp . /OCN
i li i , (25)
3s± 1 - pj ds±
3p, 1 — p. dp.
~±= \ —- (j = 1,2). (26)
8x. 1 - p. dx.
J i J
The optimality criteria (22) for the margins of different kinds of
government expenditures become,
-Z (p ) ^ —i = 1 (j = 1,2). (27)
• J- . ^
Note that [(1 - p.)/(l - P~? ) ] is the probability of survival if the indi-
vidual does not die from cause j .
We can substitute (25) into the criterion (20) , which measures the will-
ingness to pay for safety by the marginal contribution of private safety
expenditures. Let,
s?(p?) = private expenditure needed to reduce probability of death
from private causes to p., (28)
so that ds?/dp° = I / (dp°±/ ds ±) . Then,
pi dsi
(29)
the right-hand side may be more meaningful to calculate since it concen-
trates on the "technological" relation between probability of death affected
by private expenditures and those expenditures.
Life-cycle analysis. This section analyzes the situation when people
are considered as living an extended period of time. For simplicity, assume
a number of people all live from 0 to T. For the i^ individual, let
17
-------
c. = consumption at time t, (30)
s, = safety expenditures at time t, (31)
U, (c. ) = utility of consumption at time t, discounted back to
time zero, (32)
PjL(t) = probability of survival to time t. (33)
Then the individual's lifetime utility is given (assuming intertempor-
ally additive utilities) by,
'oTpl(t) Uit(cit) dt>
and therefore a Pareto analysis seeks to maximize
1 Wi C Pi(t) Uit(clt) dt' (34)
There are two kinds of constraints: those on production or the availa-
bility of resources in general and those on the relation between probability
of survival and resources devoted to safety.
With regard to the first, we make the following simple assumption. Each
individual is capable of producing something at any given age; product at any
time can be converted to product at a later time at a growth rate r. There
may be an initial accumulation of assets at time 0. We can therefore as
usual discount all resources and expenditures back to time 0 at a rate r; but
it must be kept in mind that both human resources and personal expenditures
must also be discounted by the probability of survival. To simplify notation,
define consumption to be the amount above and beyond the human resources from
the same individual at any time t; since the utility function has a time sub--
script, this adjustment can be absorbed into the form of the utility function.
Of course, c. may, and usually will, be negative from some values of t, with
this definition. Also define,
x = public safety expenditures. (35)
Then total expected expenditures (net of human resources) at time t are,
Xt+iPi(t) (cit+sit>>
and therefore the resource constraint is,
/0T e~rt [xt + £ P± (c±t + s±t)] dt = constant, (36)
the constant being the initial accumulation of assets.
18
-------
We can construct a simple model of the interrelation between safety ex-
penditures and the probability of survival as follows. The risks at differ-
ent moments of time are assumed independent; the probability of death in any
small period of time is a function of public and private safety expenditures
during that period. More explicitly, by the usual laws of probability theory,
Pi(t + h) = P.(t) Prob (survival to t + h| survival to t)
= P.(t) [1 - Prob (death between t and t 4- h| survival to t)].
This last probability is assumed to depend only on safety expenditures be-
tween t and t + h (for given t, i.e., the effect of age on probability of
death is not excluded). Let P. (t + hXt) be the probability of death between
t and t + h, given survival to t. Then taking logarithms of both sides, sub-
tracting x log P. (t) from both sides, dividing through by h, and letting h
approach zero yields,
P./P. = -lira [P (t + h|t]/h = - p (s.. , x ), say. (37)
h 0 it ic t
From the assumption, p does depend only on safety expenditures in the
interval (t, t + h) or h vanishingly small, hence only on expenditures at
time t.
The maximization of (34) is subject to the constraints (36) and (37),
We apply the Hamiltonian (Pontryagin) method. The constraint (36) is a sin-
pie overall resource constraint (frequently called "isoperimetric"
constraint in the calculus of variations), so that the dual variable associ-
ated with it is a constant over time: call it X. The constraints (37)
(one for each individual) are differential equations; the associated dual
variables are functions of time, say V.(t), and each satisfies a differen-
tial equation to be given below.
Form the so-called Hamiltonian obtained by adding the maximand to the
constraints, each weighted by the corresponding dual variable.
H - i Wi /oT Pi Ui(cit) + Xe"rt[xi + I VCit + Sit)]
- k pit(sit' v pi- (38)
(To obtain the last term, the equations (37) are written in the standard form,
P = - p. P..) Then the optimum is characterized as follows: For each t,
the flow variables, c , s. , and x are chosen to maximize H. They are thus
19
-------
determined in terms of the state variables P. and the dual variables. The
dual variables P^ (t) have to satisfy the differential equations,
PI = - H/ P1. (39)
Since the payoff is clearly not influenced by the values of P at time T,
it can be shown that the corresponding dual variables must be zero there (the
dual variable can be interpreted as the shadow price of the corresponding
state variable, i.e., the increase in payoff associated with an infinitesimal
increase in the value of P.).
y±(T) = 0 (40)
Of course, by definition,
P±(0) = 1. (41)
Hence, the problem is to solve the differential equations (37) and (39) with
the (awkward) boundary conditions (40) and (41). As the problem is set up,
X is still undetermined and must be chosen to make sure that (36) is satis-
fied.
V
We first maximize H with respect to the flow variables, c , s , and
wi pi uit + Xe~rt pi = °> (42)
Xe"rt P± = y± OPl/3slt) P±, (43)
8e~rt = Ep± Op1/3xt;) P±. (44)
Note that we can cancel P in (42) and (43).
Wi Uit = -Xe~rt» <45>
U± (9pi/8slt) = Ae"rt . (46)
The most striking implication comes from (46) in conjunction with (40).
At t = T, the right-hand side is finite (actually, negative equation from
(45) shows) while y.(T) = 0. Hence, for t = T, it must be that 3p /8s. is
infinite. If, as before, we solve for s. as a function of p, (holding
public expenditures on safety constant), we see that
9slt/8PiT = 0, (47)
20
-------
i.e., the willingness to pay for additional safety goes to zero at the end
of one's life. By continuity, we can argue that the willingness to pay will
decrease for old age, though I have not so far been able to prove that it is
a monotonic decreasing function of time. (To be sure, the assumptions made
so far are very general; but even experiments with much more specific func-
tions have not yet led to this conclusion.)
In (44), we can divide through by Xe , and then, in the i term on
the right-hand side, replace it by the left-hand side of (46) to derive the
same result as implied by (18) and (20):
I Osit/3Pi) (3P±/3x) = 1. (48)
Hence, the static results on the relative margins of public and private safety
expenditures remain valid. Further study of the life-cycle analysis may yield
stronger results.
For completeness, the differential equation (39) can be stated in
explicit form as:
P. = - w± U.t - Ae-rt (c±t + s.t) + P. p.. (49)
This still involves the welfare weight, w., which we seek to eliminate.
Therefore, manipulate this equation further, as follows. First, replace
Xe~rt in (49) from (45). Then define,
v = y /w •
i V i'
and recall that w. is a constant over time. Then divide through in (49) by
w..
Ci = - Uit + Ult (cit + Sit) + \ Pi
= [uit-uit (cit + sit)] + ViV (50)
The bracketed term is a sort of consumer's surplus, hence this equation
appears to have some meaning, as yet undefined.
Limitations of the Pareto criterion. Although the preceding analysis is
obviously quite powerful, this approach has some significant limitations.
First, if prices are not specified, it is not clear at what point the economy
will equilibrate. Also, it is impossible to distinguish, for purposes of
evaluating the level of social welfare, between various points of efficient
equilibrium (i.e., the contract curve). Second, the possible efficient equi-
libria change as endowments change. For example, if the original endowment
of some individuals is increased by a certain amount, and a like quantity is
subtracted from other individuals, the set of efficient equilibria may be
altered substantially.
21
-------
A serious, but not fatal, deficiency of the Pareto criterion for evalua-
ting social decisions involving low probability, high consequence occurrences
is that it changes as knowledge and expectations change. If knowledge
changes, expectations may change. Once expectations change the expected
utility derived from any good or bundle of goods may also change. Assuming
that individuals in the economy are risk-averse, as is implied by convex (to
the origin) indifference curves, the preference orderings for any given set
of expectations will be concave. However, if knowledge and expectations
change, individuals will reevaluate the satisfaction derived from various
goods. These revised expectations will change the utility associated with
each consumption bundle and hence change the contract curve. Thus, at any
particular point in time, an evaluation of a social decision with uncer-
tainty might conclude that a decision is correct. Yet with additional infor-
mation, obtained only with the passage of time, the conclusion might be that
the decision is incorrect. If decisions are reversible, then this limitation
of the Pareto criterion is not overwhelming. However, if decisions are ir-
reversible, it may be that any original efficient equilibrium would impose
substantial welfare losses over time.
There are other defects of the Pareto criterion. For example, it runs
into conceptual as well as measurement difficulties if people attach value
to ideologies or processes as well as to conventional commodities. It loses
some of its appeal if we introduce certain plausible interdependencies among
utility functions — for example, degrees of malevolence, jealousy, or
tastes for interfering with each other on the one hand, and on the other,
empathy for others. 2j Also, to approve this economic efficiency criterion
without any reservations whatsoever requires a particular value judgment
which many are unwilling to make. We simply do not believe it is right to
permit sadist-masochist exchanges in which one person loses an eye or is
assisted to commit suicide or sells himself into slavery. Moreover,
Scitovsky re-raises troublesome questions about rationality. 3/ Finally, for
those who believe Pareto optimality (loosely speaking, aggregate real in-
come) is linked with welfare or happiness, several people including Scitov-
sky have presented arguments that reinforce one's intuitive doubts about
strength of the long-run correlation.
Social Decisions and Processes
This section surveys the major components of the contract-based theories
developed by Rawls and Buchanan. Their primary evaluative criterion for
social decision-making is that, if the process of making the decision is just,
then the decision, whatever it may be, is the socially correct one. In other
words, it is not necessary to evaluate the "optimality" of the decision it-
self. The decision is believed to be optimal once it is determined whether
the process is just. This complete refutation of "the ends justify the
means" seems particularly appropriate for consideration in the post-
Watergate morality.
Rawls' Theory of Justice. Rawls' theory is based on two principles of
justice: 4/
22
-------
First Principle — Each person is to have an equal right to the
most extensive total system of basic liberties compatible with a similar
system of liberty for all.
Second Principle — Social and economic inequalities are to be ar-
ranged so that they both are:
(a) to the greatest benefit of the least advantaged, consistent with
the just savings principle, 5/ and
(b) attached to offices and positions open to all under conditions
of fair equality of opportunity.
Rawls derives these "principles" from his definition of the "original
position." The original position is populated by individuals who are all
motivated by self-interest. Further, individuals are behind a "veil of ig-
norance" and do not know their endowment or status in any future state of
the world. Rawls argues that combining the veil of ignorance and motivations
of self-interest assures adoption of these two principles of justice, and
will be ranked in lexical order. That is, no diminution of the first prin-
ciple is allowed to obtain the second.
Criticism of Rawls' theory of justice are extensive. _6/ Significant
limitations of Rawls' theory of justice include the artificial use of time,
as required by the veil of ignorance, and the implied risk aversion or ignor-
ance associated with adoption of the second principle of justice. The second
principle of justice implies that an action is good if it does not decrease
expectations of the most disadvantaged member of society. For example, the
optimal social decision is the one that maximizes income of the poorest
(minimum) individual. This "maximin" principle will be chosen if those in
the original positions are infinitely risk averse. ]J
If it could be shown that adoption of Rawls' second principle is only
consistent with infinite risk aversion, this would be a significant weakening
of Rawls' argument for acceptance of the two principles of justice. How-
ever, Rawls claims that infinite risk aversion is simply consistent with the
assumptions he makes concerning the original position, J3/ Rawls requires
that individuals in the original position prefer the "well ordered" society.
One component of a well ordered society is that any contract entered in the
original position be one that each individual in the original position thinks
he can "live with." Since, for Rawls, each individual in the original posi-
tion is representative of himself and his future generations, this requires
that each individual in the original position be able to "live with" the con-
tract no matter what the future brings. If each individual in the original
position could not live in a society (maintain his contract) if he or his
future generations were sufficiently disadvantaged, then adoption of the
maximin principle is consistent with the desire for a well ordered society,
since a well ordered society requires that each individual be willing to
abide by the contract developed in the original position.
Rawls tries to include time by having each person in the original posi-
tion represent himself and all future generations. In this manner, Rawls
attempts to derive principles of justice that will not change with the pas-
sage of time. This particular way of incorporating time means that Rawls1
23
-------
model remains static, and changes in knowledge cannot change the two princi-
ples of justice. At the same time, nothing seems to be repugnant about
changes in the "principles of justice" over time since knowledge changes,
provided the process of determining these changes is acceptable to society.
Another substantial criticism of Rawls' theory is that it implies a
certain uniformity in the preferences of individuals in the original position.
There cannot be substantial differences in their conceptions of justice and
future possible states of the world or their perceptions of their effect on
future possible states of the world.
Rawls argues for selection of the two principles of justice by contrast-
ing the choice between his two principles and other possible rules such as
total utility maximization, average utility maximization, etc. These alter-
natives, which are the choices considered by Rawls to be available to indivi-
duals in the original position, are not very dense. Other alternatives not
unambiguously inferior to the two principles of justice may also exist. If
some ambiguity does exist, then the only remaining way to maintain the choice
of Rawls1 two principles of justice is to argue on the basis of their "moral"
worth, indicating that "right" men would necessarily prefer the two princi-
ples of justice. However, imposing some form of outside morality on their
choices would upset Rawls' structure. Such an argument contradicts Rawls'
emphasis that the two principles of justice are the only logical choices,
given certain minimal axioms concerning individuals in the original position.
If one accepts Rawls' theory of justice, then the criteria for evalua-
ting a social decision concerning a low probability, high consequence occur-
rence are relatively straightforward. First, does the social decision-
making process maintain the first principle of justice? Or, does everyone
have equal opportunity with respect to the social institution charged with
making the decision, and are there other appropriate safeguards for this
equality of opportunity during the decision-making process? Once this prin-
ciple is fulfilled the question then becomes whether the social decision-
making process and the decision are such that the most disadvantaged member
of society does not suffer lowered expectations as a result of the decision.
Obviously, the most disadvantaged member might change after a decision
is made. Rawls argues that this is an exceptional case due to what he terms
a "chain connection". 9/ If the most disadvantaged member changes, Rawls1
criteria imply that the expectations of the most disadvantaged, both before
and after, must not decrease.
Evaluation of a social institution and process relative to the first
principle of justice is qualitative. You either fulfill the first principle
or you violate it. There are no degrees of fulfillment given the lexical
order of the first and second principles. Measurement problems, which hinder
application of either the average utility criterion or the Pareto criterion,
remain in identifying the most disadvantaged member of society as well as
determining whether that person's expectations have increased or decreased.
Buchanan— The Limits of Liberty. Whereas Rawls effectively created a
world where conflicts between individual preferences could not occur — the
24
-------
original position — Buchanan accepts the conflicting nature of individual's
desires, and tries to determine whether these conflicts and their resolution
imply that the state will grow beyond its "legitimate" limit. 10/ Buchanan
begins from the basis of anarchy and describes how social institutions would
evolve due to a recognition that a higher degree of Pareto optimality can be
obtained through cooperation — for example, a mutual agreement not to attack
each other in a two-person society. Or, in a multi-person society, some sub-
set of its members might agree to a mutual protection of their property, the
endowment of which is a result of "anarchistic equilibrium."
Buchanan accepts the continuing revision of social contracts, in con-
trast to Rawls. As long as government does not exceed its legitimate limits,
where the legitimate limits are defined by the Pareto criterion, any indivi-
dually and freely chosen social contract is optimal since to conclude other-
wise would imply that individuals were behaving irrationally. Once govern-
ment goes beyond its limits, Buchanan offers no criteria for choosing among
these inferior positions.
Buchanan's efforts seem more positive than normative. With little dis-
cussion, he accepts the Pareto criterion as the mechanism for defining suit-
able limits of government. Once these limits are defined, he demonstrates
that individuals, maximizing their self-satisfaction, will create social
institutions. In a world where social decision-making requires limited re-
sources, these institutions will cross the acceptable limits defined by the
Pareto criterion. Buchanan's primary focus is on showing why societies that
begin with relatively few social institutions (in other words, the United
States) inevitably end up with too many social institutions, and the Pareto
efficiency of the society is hindered. In the context of this discussion,
though, Buchanan's views provide an effective means of appraising criteria
derived from Rawls' theory.
Summary
This chapter has discussed rather different criteria for evaluating
social decisions. The next chapter will argue for a "mixed" approach to
evaluating social decisions involving low probability, high consequence
occurrences. The mixed approach is composed of (1) the dominant ethic of
present day evaluations: an act or a process is good if its consequences
are independently demonstrated to be good, and (2) the contract-based theory
and principle of ethics implied by Rawls and Buchanan: the good is that
which emerges from a proper procedure.
The argument for this mixed approach is based in substantial part upon
the lack of knowledge and associated uncertainty that characterizes social
decisions involving low probability, high consequence occurrences. Although
an evaluation can estimate the expected net benefits of a decision, this
present estimate of net benefits is obviously not necessarily equal to the
actual benefits that will be incurred. If possible actual outcomes are suf-
ficiently diverse — if there is a great deal of uncertainty — then conflic-
ting estimates of the expected net benefits are likely. It may be impossible
to convince each member of society that any particular estimate, and decision,
25
-------
is superior to all alternatives. Thus, in addition to generating good deci-
sions, it is also necessary that the social decision-making process lessen or
resolve conflicts, in order that the final decision will be accepted and,
therefore, will be implementable.
Another important reason for the mixed approach recommended here is that
social decisions involving low probability, high consequence occurrences im-
pose risks upon members of present and future generations. Imposition of
these risks appropriates a portion of the property rights of individuals in
present and future generations, since they are no longer able to choose to
reduce the risk they bear below the new, higher minimum established by the
social decision. Definition of property rights has historically been charged
to social institutions. Validity of these definitions has been evaluated in
terms of the quality of the process used to develop them as well as the
justice of the definitions. For these reasons a "proper process" is
important.
Notes to Chapter 3
1. See Scott Gordon, "The New Contractarians". Journal of Political Econ-
omy (June, 1976), pp. 573-590.
2. Feeling for others that transcends simple market values manifests itself
in many ways and has long been recognized as a force by both economists
and noneconomists alike. A familiar example of a noneconomic statement
of such feelings is the following: "No man is an Hand, intire of it
selfe; every man is a peece of the Continent, a part of the maine; if a
Clod bee washed away by the Sea, Europe is the lesse, as well as if a
Promontorie were, as well as if a Manner of thy friends or of thine owne
were; any mans death diminishes me, because I am involved in Mankinde;
And therefore never send to know for whom the bell tolls; It tolls for
thee." John Donne.
3. Tibor Scitovsky, The Joyless Economy (New York: Oxford University
Press, 1976).
4. John Pawls, A Theory of Justice {Cambridge, Mass.: Belkamp, 1971),
p. 302.
5. The just savings principle requires that each member of the original po-
sition represent family lines, with ties of sentiment between successive
generations. These ties are then the basis for determining the criteria
for justice between generations, for example, a specified minimum rate
of savings. See Rawls, A Theory of Justice, pp. 291-93.
6. See Gordon, "The Contractarians"; Sidney S. Alexander, "Social Evalua-
tion Through National Choice," Quarterly Journal of Economics^ (November
1, 1974), pp. 625-32, and R. A. Musgrave, "Maximum, Uncertainty, and the
Leisure Tradeoffs," Quarterly Journal of Economics (November 1974),
pp. 625-32.
26
-------
7. Musgrave, "Maximum, Uncertainty and Leisure Tradeoff".
8. John Rawls, "Reply," Quarterly Journal of Economics (November 1974),
pp. 633-55.
9. Rawls, A Theory of Justice, pp. 81-83.
10. James M. Buchanan, The Limits of Liberty; Between Anarchy and Leviathan
(Chicago: University of Chicago Press, 1975).
27
-------
4^ Criteria For Evaluating Social_Decision Processes
Characteristics that distinguish environmental risk problems were de-
scribed in Chapter 2. Many of these characteristics can give rise to severe
contention among various members of society.
Limited knowledge, for instance, may make any general agreement on ex-
pected benefits and costs impossible. Because the estimated probabilities
are so low, and typically have no empirical basis, there is often a wide
range of estimates, even among knowledgeable and informed scientists. This
range of uncertainty in turn leads to large differences concerning expected
costs. For example, consider a high consequence outcome with a cost of $500
billion ($5 x 10 ) if it occurs. Also, assume that the plausible low proba-
bility estimates range from 10 to 10 . Since the consequence is so
costly if it occurs, two technical experts who believe that the appropriate
low probability estimate is at either extreme of the "possible" range have
implied expected costs ranging from $5 (trivial) to $500,000 (substantial).
Many situations that give rise to environmental risk do not represent
technological breakthroughs with extremely high economic payoffs. Rather,
the expected net benefits are often in a range comparable to other invest-
ments that do not involve any known environmental risk. This makes uncer-
tainty concerning the high consequence outcome critical to evaluation of the
project. To illustrate, assume that the certain costs of the project are
$1,000,000 and certain revenues are $1,100,000. Known revenues and costs
imply that the project is beneficial. However, if cost of the consequence,
in case a bad event occurs, is agreed to be $500 billion, the lowest low
probability (10 ) implies pursuing the project while the highest low proba-
-6
bility (10 ) implies a substantial expected loss from pursuing the project.
In this example an optimal decision depends critically upon this expected
cost item.
If uncertainty concerning the probability estimate is coupled with un-
certainty about the expected cost of a bad occurrence, the potential for
significant differences of opinion increases. For many environmental risk
situations, some people will feel they are "losers" no matter what decision
is made.
Yet another source of potential discord and contention arises from the
fact that decisions involving low probability, high consequence occurrences
often have implications for future generations. It is impossible to account
perfectly for the preferences of future generations, since their prefer-
29
-------
ences are not known, but a well-ordered society (to use Rawls' phrase) re-
quires some explicit recognition of these future generations and their pos-
sible state of affairs.
Economic efficiency is, of course, also important. Under certain condi-
tions, a free market structure permits the greatest value to be obtained from
a given set of resources. A benefit-risk-cost analysis applies values ob-
served in the market to develop an estimate of the net return associated with
a project. A "complete" benefit-risk-cost analysis will in theory enable a
social decision-maker to make the optimal choice. Obviously a benefit-risk-
cost analysis is never complete. In many instances, though, including many
problems of classical environmental pollution, it provides a close enough
approximation to serve as the principle basis for the decision process. How-
ever, characteristics of low probability, high consequence occurrences, as
outlined in Chapter 2, coupled with difficulties of applying the Pareto cri-
terion, described in Chapter 3, make it impossible to maintain that a benefit-
risk-cost analysis can imbed an acceptable consideration of future genera-
tions as well as resolve contention over widely varying estimates by "ex-
perts" on what are often highly charged emotional issues.
Economically efficient solutions that seriously undermine the social
order are not socially optimal in a broader perspective. Economic efficiency
must be pursued within the constraint of social acceptability. There are too
many caveats to rely solely on a benefit-risk-cost analysis, or, for that
matter, Rawls' znaximin criterion. A process is required that will make an
incorrect decision with only a low probability while minimizing contention.
Such an approach will necessarily include components of a "good" benefit-
risk-cost analysis. It will also include some components of maximin and
other criteria required to convey the belief that affected individuals have
had their preferences and desires considered. Thus, a proper evaluation of
social decision processes involving low probability, high consequence occur-
rences requires a mixed set of criteria.
Outcome and Process Evaluation
In order to develop a set of criteria, we have combined requirements for
economic efficiency with requirements for an acceptable social decision pro-
cess. These criteria can be used to evaluate legislative hearings and delib-
erations, proceedings of the National Academy of Science, hearings and pro-
ceedings held under the Administrative Practices Act, legal proceedings in
the courts, and proceedings before any new tribunal such as a Science Court,
should it ever be established.
The applicability of existing institutions for handling problems of en-
vironmental risk are being called increasingly into question. It is widely
recognized, for instance, that the legal system pursues the truth within a
very special framework. The objective is to limit the chance of a false con-
viction, on grounds that it is better to free many guilty people rather than
convict one innocent person. Under these circumstances the truth is occa-
sionally secondary to the process •— for instance, evidence obtained ille-
gally is not admissable, regardless of how firm it is or how much light it
30
-------
might shed on the guilt of the accused. This does not mean that the legal
system does not search for the truth, but simply that truth is sought only
within constraints imposed by the process. Some characteristics of the pro-
cess are valuable enough to warrant decreased accuracy. For example, the
process guards rights that society has assigned to individual liberty. Also,
it explicitly considers the bases of contention so that opposing parties are
generally willing to accept the decision.
The legal system evolved as a means of resolving past occurrences,
either crimes that had been committed or civil disputes involving some prior
incident such as breach of contract or injury of one party by another. While
the legal process can perhaps be improved marginally, it seems generally suf-
ficient for its present purposes. If it makes a mistake, the outcome does
not involve unusually high cost to society. As long as the legal system
maintains a certain rate of accuracy, no argument is seriously accepted that
the legal system should be changed because it does not always result in an
accurate decision.
Throughout its long evolution, however, the legal system was never in-
tended to be a future-oriented decision process, and many are beginning to
question whether the procedures and biases which are so deeply rooted in the
established legal process represent the best choices for making future-
oriented decisions. As noted previously, there are two types of error in a
criminal case: (1) falsely convicting an innocent person, and (2) failing to
convict a guilty person. Decisions involving environmental risk also have
two types of possible error: (1) falsely predicting a catastrophic outcome
from a relatively safe or harmless situation, and (2) failing to predict a
catastrophic outcome. The criminal process seeks to minimize the first type
of error, and many feel that the social decision process should be designed
to minimize the second type of error. That is, it should err in the direc-
tion of rejecting projects that are not, in fact, all that risky in order to
avoid the mistake of accepting projects that are, in fact, risky. In other
words, mistakes should have a very low probability of occurrence. Since
there may be disagreement over the correct decision, an acceptable process is
necessary in order that the decision, whatever it may be, is accepted.
Table 1 summarizes the primary source for each of the evaluation criteria
developed for this project. As indicated there, the economic, maximin, and
contract approaches have all been used to derive these criteria. The de-
scription of each criterion also includes some discussion of the source.
Evaluation Criteria
The criteria that should be fulfilled by a process making social deci-
sions involving low probability, high consequence occurrences include:
1. Are parties who incur the present costs adequately represented?
2. Are the parties who incur the present benefits represented? Criteria
1 and 2 insure the present parties to the decision come together. These cri-
teria are necessary for an optimal trading arrangement as defined by
Pareto. They are also necessary requirements for acceptance of a decision
31
-------
TABLE 1
Proposed Evaluation Criteria & Their Derivation
Economic
Criterion (Pareto)
1. X
2. x
3.
4. x
5.
6.
a.
Maximin Contract
(Rawls) (Buchanan)
x x
x x
x
x x
x
x x
b. x
c.
x
d. x
e.
x
£. x
7.
x x
8. x
32
-------
that comes out of the process, particularly if the distribution of benefits
differs from the distribution of costs.
3. Are interests of the future generations suitably considered? Social
decisions that impose benefits, costs, and risks over a long period of time
make it necessary to consider future generations, yet consideration is neces-
sarily limited by our lack of knowledge concerning future states of the world
and future preferencs. At a minimum, however, it is necessary to describe,
as accurately as possible, possible impacts on future generations. Once this
description is obtained, members of society can, if they choose, attempt to
represent both themselves and future generations as in Rawls' original posi-
tion. Criterion (3) goes beyond the usual role of the social discount rate
in the benefit-risk-cost analysis. Interdependence of risks over time,
coupled with considerations of irreversibility and uncertainty concerning
future preferences generally make it impossible to consider adequately
future generations through a social discount rate.
4. Are distributional aspects considered systematically? Distribu-
tional aspects of the social decision may be significant. If they are, com-
pensation (through the tax system, perhaps) is necessary. At a minimum,
groups who gain or lose should be identified for each alternative decision.
This criterion is of importance to acceptance more than optimality. A social
decision that makes the rich better off without imposing any costs upon the
poor is an economically efficient action. Acceptance of this social decision
may nevertheless require that some of the additional income be redistributed
to the lower classes, in order to maintain stability of society, especially
if there is some uncertainty about the distributional aspects. Finally,
Rawls1 maximin criterion requires that the least advantaged (e.g., poorest)
not be injured by the decision.
5. Are social decision makers "unbiased" and open? In a free market
prices are unbiased in that they are not influenced by any one individual. A
particular preference or opinion influences prices only if held by signifi-
cant numbers of individuals. A similar property is desired of decision-
makers since they are effectively allocating resources among competing claims.
This is the "unbiased" property of the decision-maker and is necessary for an
optimal decision. Also, the decision is more likely to be accepted by losers
if the decision-makers are perceived to be unbiased. The other characteristic
is that the decision-maker be "open". The decision process for low probabil-
ity, high consequence occurrences must not preclude intermediate decisions.
If the process indicates sufficient uncertainty, the decision-maker must be
able to make limited choices, i.e., to keep his options open. If uncertain as
to the impact of a social decision, it may be wise to incur a higher expected
present cost by limiting the program, while more accurate information is
developed through experimentation and observation of the limited program.
6. Development of knowledge concerning the effect of possible alterna-
tive decisions:
a. Do views that appear to come from reasonable sources receive — or
appear to receive — "equal time and treatment"?
b. Are all significant differences considering quality of knowledge
included?
33
-------
c. Are gaps in knowledge identified?
d. Are procedures for collecting knowledge of the area efficient?
e. Are plausible consequences for the worst contingency spelled out?
f. Are alternatives suitably considered? That is, does the process
explicitly recognize the uncertainty and lack of knowledge about
pertinent alternatives or possibilities?
Criterion (6) is the means of evaluating the effectiveness of the process at
eliciting the available knowledge, defining the areas of uncertainty, and de-
fining the possible alternative decisions. Obviously, it is necessary that a
complete benefit-risk-cost analysis fulfill this criterion. In order to make
an optimal choice, in any sense, it is necessary that we determine what is
known and not known about possible states of the world. Once this knowledge
is obtained, it is necessary to define the possible alternative decisions
since optimality of any one decision or mix of decisions will be dependent
upon our knowledge.
7. Is information given to individuals who are at risk? Those who will
experience an increase in their level of involuntary risk should obviously be
informed, since their property rights are being altered. This criterion for
writing a new contract is required by both Rawls and Buchanan. In order to
inform these individuals it is necessary that knowledge developed by the pro-
cess be translated into "lay" terms, including succinct presentation of the
state of disagreement, especially if scientists reach quite different judg-
ments. Fulfillment of this criterion should increase acceptance of the deci-
sion. While not required by the criterion of Pareto efficiency, in practice
it might also increase the probability of an optimal decision from the
process.
8. Can a decision change if significant new information is obtained?
This final criterion is particularly sensitive since frequent change is un-
desirable due to possibly high transition costs. Yet, unlike legal proceed-
ings, the decision process must not rely too heavily on historical precedent.
This criterion recognizes the dynamic implications of many social decisions
involving low probability, high consequence occurrences. When an individual
decides upon a path of action over time, he is free to change his path if his
knowledge and expectations change. However, the individual will not change
his actions with every change in his knowledge and expectations. Some changes
will not be sufficient to warrant incurring the costs of change. Optimality
of the social decision over time requires that the decision process must have
an element that defines circumstances for reconsidering a decision.
Obviously, each of these criteria is subject to dispute. Also, it is
only possible to formulate such criteria in a finite number of pages by allud-
ing to words like good, adequate and appropriate. The criteria are derived
from a desire to see correct social decisions made and accepted. Some of
these criteria are verbal adaptations of the steps necessary in a good
benefit-risk-cost analysis. A verbal adaptation is necessary since our lack
of knowledge concerning social decisions involving low probability, high
consequence occurrences insures that there will be disagreement about some of
the components. Other criteria are based primarily on a desire that decisions
34
-------
of a "good" process be acceptable. Thus, the process has value in terms of
both optimality of its decisions and acceptability by the affected indivi-
duals .
These criteria avoid both extremes: Pareto efficiency implies that a
process is good if it gives consequences that are independently demonstrated
to be good. The ethics implied by Rawls and Buchanan, at the other extreme,
are that the good is that which emerges from a proper procedure. If a process
meets the eight proposed criteria, then it fulfills both principles of ethics.
Some of the criteria insure that the decision will be optimal except in a low
number of cases. Other criteria insure that the process and its decision will
be acceptable to affected individuals because the process is, in certain spe-
cified senses, proper.
Conclusion
The problems facing a benefit-risk-cost analysis of low probability,
high consequence occurrence are so many as to make any analysis subject to
wide dispute. Certainly, benefit-risk-cost analyses of pesticides or nuclear
power are always arguable. Controversy seems to center on whether we are
willing to assume present risks and whether we are willing to impose risks on
future generations.
If a decision tests the limits of our knowledge sufficiently, or if the
risk is severe enough, we will not accept the market process or its surrogate
(benefit-risk-cost analysis) without other procedures associated with social
institutions. The reasons have been discussed throughout this paper. The
conclusion is that a process for making social decisions involving low
probability, high consequence occurrences should fulfill certain criteria
that imply a high probability of an optimal decision as well as acceptance of
the decision by the affected parties.
35
-------
References for Part I
Alexander, Sydney S., "Social Evaluation through National Choice," Quarterly
Journal of Economics, Vol. 88 (November 1975) pp. 625-632.
Arrow, Kenneth J., Essays in the Theory of Risk Bearing, Chicago, Illinois:
Markham Publishing, 1971.
, Social Choice and Individual Values, New York: Wiley, 1951.
, and Robert C. Lind, "Uncertainty and the Evaluation of Public Invest-
ment Decisions," American Economic Review (June 1970) pp. 364-78.
Buchanan, James M., The Limits of Liberty: Between Anarchy and Leviathan,
Chicago: University of Chicago Press, 1975.
, and Gordon Tullock, Che Calculus of Consent, Ann Arbor: University
of Michigan Press, 19627
Friedman, Milton and L. J. Savage, "Utility Analysis and Choices Involving
Risks," Journal of Political Economy, Vol. 56 (August 1948).
Gordon, Scott, "The New Contractarians," Journal of Political Economy,
Vol. 84 (June 1976) pp. 573-90.
Green, Jerry R., "Information, Efficiency and Equilibrium," Harvard Institute
of Economic Research, Discussion Paper No. 284 (March 1973).
Harberger, Arnold, "Professor Arrow and the Discount Rate" in Project
Evaluation. Chicago: University of Chicago Press, 1976.
Kneese, A.Ian B. , "The Faustian Bargain," _Respur_ces, No. 44, Washington, D.C.:
Resources for the Future, Inc. (September 1973) pp. 1-3.
Levy, David, "Economic Theory and Ethical Judgment," unpublished (March 1975),
Musgrave, R. A., "Maximum, Uncertainty, and the Leisure Tradeoff," Quarterly
Journal of Economics, Vol. 88 (November 1974) pp. 625-32.
Nozick, Robert, Anarchy, State and Utopia, New York: Basic, 1974.
Rawls, John, A Theory of Justice, Cambridge, Mass.: Belkamp, 1971.
37
-------
, "Reply," Quarterly Journal of Economics. Vol. 88 (November 1974)
pp. 633-55.
Scitovsky, Tibor, The Joyless Economy, New York: Oxford University Press,
1976.
Shackle, G. L. S. , Epistometics and Economics, Cambridge: Cambridge Univer-
sity Press, 1972.
Weinberg, Alan M. , "Social Institutions and Nuclear Energy," Science,
Vol. 177 (July 7, 1972) pp. 31-34.
38
-------
PART II
CASE STUDY OF CHLORDANE/HEPTACHLOR
39
-------
5. Introduction
This case study is concerned with social decisions about situations
involving low-probability, high-consequence events. When they occur, such
events are usually referred to as "disasters" or "catastrophes." Low-
probability, high-consequence events are sometimes caused by natural occur-
rences, such as earthquakes, floods or tidal waves, but history has also wit-
nessed high-consequence events not caused by nature. A few well-known
examples are the Chicago Fire, the Texas City explosion, the underwater oil
blowout in the Santa Barbara channel, and the numerous birth defects caused
by thalidomide. Some of these disasters might have been avoided had someone,
or some authority, taken timely action.
Any modern, high-technology society is susceptible to a wide variety of
low-probability, high-consequence events: (1) air crashes of fully-loaded
giant aircraft; (2) the release of large quantities of poisonous gas near a
population center while being transported; (3) major explosions or oil-spills
from giant supertankers while in harbors of major cities; or (4) nuclear
reactor accidents that result in large-scale release of radioactive materials.
Government decisions have been and are being made in all these areas, but
this does not necessarily mean the decisions are correct, or that the social
decision process used to determine our choices is as good as it could be.
This is a case study of the existing collective decision making process
in one area of environmental concern: namely, low-probability high-
consequence events that may arise from the use of pesticides. All chemical
pesticides contain biologically active ingredients, and are by design, toxic
to some living organisms. It has long been recognized that this toxicity can
pose an inherent threat to humans who are overexposed to the chemical. Given
a high enough level of exposure, people will suffer adverse toxic effects
with high probability.
The key concern of this case study is not with the risk of direct
toxicity, but with a different kind of risk posed by chemical pesticides —
the possibility that the introduction and use of some pesticides may cause
widespread cancer, thereby shortening the lives of thousands, or even tens of
thousands of people. On the basis of what is known about the causes of can-
cer, the probability that chemical pesticides might be the direct cause of so
many deaths does not appear to be high. At the same time, it is now known
that residues from some pesticides do find their way into our food supply.
If these residues can and possibly do cause cancer in some people, then we
are faced with a low-probability, high-consequence situation. The effects of
carcinogenic pesticide residues may not be as dramatic as other disasters and
catastrophes, but the results are equally serious, if not more so. This case
study is therefore focused on this situation, the problems that it raises,
41
-------
and the decision making procedures and public policies appropriate to control-
ling it.
The first pesticide to be widely questioned and challenged because of
its environmental effects was DDT. Attacks against DDT increased in inten-
sity during the 1960's and, after a widely-publicized proceeding, culminated
in cancellation of registration on June 14, 1972. I/ The possibility of DDT
residue finding its way into food, and the possibility that such residue
might be carcinogenic, were but two of the many adverse effects ascribed to
DDT. It was alleged to have a large number of adverse environmental effects,
including harm to fish and wildlife, especially birds. The findings of fact,
both at the time of the cancellation hearing and in a subsequent follow-up
study by EPA, 2J indicate that DDT did indeed cause a large number of environ-
mental problems that justified the cancellation of its registration.
By the time the cancellation order was finally issued and implemented,
it was widely known that the use of DDT was declining because many pests were
developing a resistance to DDT. Because the cancellation decision involved
so many factors, and because use might well have continued to drop off even
without regulatory action, DDT does not provide a clear-cut example for deal-
ing with risks posed by pesticides that are possibly carcinogenic. _3/ The
facts and evidence developed at the DDT hearing did, however, indirectly
raise important questions concerning other pesticides, especially chlorinated
hydrocarbons such as aldrin, dieldrin, chlordane and heptachlor. All of these
pesticides have now been the subject of adversary regulatory proceedings. We
thus have available for review extensive, documented public records of the
current decision-making process in operation.
The study deals with the two pesticides chlordane and heptachlor, be-
cause (a) they have been the focus of an extensive regulatory decision-
making process by the Environmental Protection Agency, and (b) the principal
basis for challenging the registration of these two pesticides was their car-
cinogenicity. Chlordane and heptachlor, as well as the other chlorinated
hydrocarbons, are effective against a broad range of pests and persist in the
environment for several years after application. These characteristics make
them highly desirable products for pest control, and have resulted in their
widespread application to many crops and millions of acres of land.
Chapter 6 surveys the development of federal pesticide regulation espe-
cially as it relates to cancer and other serious long-term side effects.
This survey includes a description of the current decision-making process as
embodied in existing legislation and the regulations of executive agencies.
At the heart of this process is a set of quasi-judicial adversary proceedings
whose basic philosophy has been unchanged since enactment of key pesticide
legislation in 1972, even though legislative amendments (1975) and new admin-
istrative regulations within the Environmental Protection Agency (1976) have
changed some of the detailed triggering and review mechanisms. 4_/
Chapter 7 takes a closer look at these adversary proceedings and their
administrative setting in the case of chlordane and heptachlor. This is a
major proceeding with at least one significant phase (suspension hearings
and decision) completed at the time this study was undertaken. The chlor-
42
-------
dane/heptachlor case represents a procedural divide: it was the last to be
initiated under the administrative rules preceding the 1976 changes, and as
such provides the best test case available for discussing differences which
the new rules are likely to make with regard to the effectiveness of the
decision-making-process.
Chapter 8 contains an extensive analysis and discussion of the theoreti-
cal issues that arise in connection with decisions involving the risk of can-
cer from pesticides. Materials from preceding chapters are used to place
these theoretical issues within the context of the current decision-making
process.
Chapter 9 examines the pesticide decision-making process in the broader
setting of various institutional arrangements designed to gather, present
and weigh scientific evidence on controversial issues. This chapter puts
the existing decision process into a perspective that spans both existing
and proposed alternative institutions.
The concluding chapter, (Chapter 10), contains an evaluation of the
current decision process used for pesticide products. The criteria for this
evaluation are those developed in Part I of this study.
Notes to Chapter 5
1. The DDT cancellation hearing lasted approximately seven months. See
U.S. Environmental Protection Agency, Environmental Facts - DDT and
the Environment (Washington, D.C.: Environmental Protection Agency
July, 1974).
2. U.S. Environmental Protection Agency. DDT: A Review of Scientific
and Economic Aspects of the Decision to Ban its Use as a Pesticide.
(Washington, D.C.: Environmental Protection Agency, July, 1975) (EPA-
540/1-75.022).
3. Carcinogenicity is not the only low-probability, high-consequence occur-
rence that can result from pesticides. Catastrophic non-cancer impacts
are also possible. For example, certain ingredients or impurities in
herbicides may destroy the viability of life sustaining support systems
or result in mass birth defects through mutagenic or teratogenic effects.
4. Detailed citations will be found in Chapter 6.
43
-------
6. Review of Federal Pesticide Regulation
This chapter traces the historical development of the legislative and
regulatory framework within which decisions concerning chlordane and hepta-
chlor have been made. The purpose of this review is to identify issues per-
tinent to decision-making about low-probability, high-consequence events,
and to put the subsequent discussion about chlordane and heptachlor into
context.
Federal pesticide regulation is viewed here as occurring in several
waves that originate at different times, and are associated with the emer-
gence and disposition of distinct issues. Each wave has several phases,
drawn out over years or decades, with different phases of separate waves
often overlapping in time. The principal phases are:
1. Legislation, as enacted by Congress. This phase in-
cludes formation of public awareness concerning issues,
which necessarily precedes legislation, together with
the political process that accompanies enactment of the
legislation.
2. Regulation, which follows the legislative phase. For
pesticides this is the registration process, including
the labelling, application and record-keeping associated
with the use of pesticides, plus the continuing infor-
mation gathering and monitoring process.
3. Adversary proceedings, including all suspension and
cancellation proceedings before administrative law
judges, procedures governing such proceedings, and
appeals to the federal courts.
Subsequent sections of this chapter review each of these phases as it
relates to the problem of low-probability, high-consequence results of pesti-
cide use. Special attention is given to the possibility that thousands, or
even tens of thousands of people may eventually die from cancer as a result
of consuming (1) chemically-treated agricultural products containing residues
of carcinogenic pesticides; or (2) other food products — such as fish —
into which such pesticide residues find their way via watersheds and the
ecological food chain.
Extensive public concern with the issue of such risks is relatively
recent. Although Congress has now enacted legislation aimed at coping with
low-probability, high-consequence events, and although it is readily apparent
45
-------
that government as a whole is taking more action than in the past, serious
concerns persist. It is not clear that the ri_ght_ kind of actions have been
taken, that the actions being taken are adequate to the problems, or that
the entire process is approaching anything that might be called gocially
o_p_t_ima 1. Numerous scientists have expressed such serious doubts and concern
about the existing process that one possible alternative, known as the
Science Court, is in advanced stages of formulation. Further discussion of
alternative decision processes is deferred until Chapters 9 and 10.
Fjgderal Legislation Dealing with Pesticides
Early History. Federal involvement in the regulation of pesticides
began in 1910 with the adoption of the Federal Insecticide Act. _!_/ This Act
was designed to prevent the manufacture, sale or transportation of adulterated
or misbranded insecticides, and was administered by the Department of Agri-
culture. This Act was intended primarily to protect farmers and other pesti-
cide purchasers who could not individually afford to analyze chemical products
in order to protect themselves from fraud. Such fraud would have reduced the
productivity of farms, as well as the income of farmers. These potential
outcomes were directly counter to the objectives and concerns of the Depart-
ment of Agriculture, so it was designated by this legislation as the agency
to oversee the prescribed regulation of pesticides. This early Act recog-
nized pesticides as toxic substances, but was chiefly aimed at providing
accurate information to purchasers, and placed almost full reliance on the
principle of cave .at emptor for the subsequent safety and protection of per-
sons using these substances. Registration of existing pesticides was not
required by this 1910 Act.
The next significant legislative action was in 1938, when Congress re-
vised the Food, Drug and Cosmetic Act. 27 Among other things, this Act
charged the Food and Drug Administration (FDA) with the responsibility for
keeping poisonous insecticides out of food products. Significantly, this law
did not require manufacturers to establish the safety of food additives in
advance of actual usage. Instead, it was left up to the FDA to discover the
use of additives, and make the tests to prove them "poisonous or deleterious"
to the satisfaction of the court when action was required to remove them from
the market. This 1938 law included provisions for setting tolerances for
poisonous ingredients, when justifiable, but the procedure was cumbersome,
time-consuming, and expensive.
Enactment of FIFRA. Following World War II, Congress enacted the
Federal Insecticide, Fungicide and Rodenticide Act (FIFRA) in 1947. 3/ This
Act extended the 1910 Federal Insecticide Act to include rodenticides and
herbicides, and required premarketing registration of products to be shipped
in interstate commerce for the first time. This Act charged that a complete
copy of the label and a statement of all claims to be made for the material
be submitted to the Secretary of Agriculture and approved before the material
could be registered for sale in interstate commerce. Like the 1910 Act, the
1947 FIFRA legislation was, above all, concerned with protecting consumers
from misbranded products. And, like the 1938 Food, Drug and Cosmetic Act, it
focused on direct toxic effects and gave no specific attention to indirect or
46
-------
long-delayed effects such as carcinogenicity. Chlordane and heptachlor were
initially registered under the 1947 FIFRA legislation. The authority to deny
registration applications was not provided until 1964. 4y
By 1951, a flood of new chemicals, together with growing concern over
the increasing use of chemicals in food products, led the House to appoint a
Select Committee to investigate the use of chemicals in food and cosmetics.
In 1952, the majority report of this Committee (known as the Delaney Com-
mittee) indicated a belief that existing laws were inadequate. After some
delay, Congress amended the Federal Food, Drug and Cosmetic Act to include
the Pesticides Chemicals Amendment in 1954. _5_/ Like previous legislation,
this law focused on the toxicity of pesticides and the procedures for deter-
mining tolerance levels. Although procedures for establishing tolerances
were administered by the Department of Health, Education and Welfare (HEW),
the Department of Agriculture was required to express an opinion as to
whether proposed tolerances were reasonable and whether residues were likely
to result from proposed patterns of use.
The Delaney Clause on Carcinogens. Public concern over chemicals in the
food supply continued to mount, and in 1958 Congress enacted the Food Addi-
tive Amendment to the Federal Food, Drug and Cosmetic Act. _6/ This amend-
ment established principles, processes, and appeals procedures for regulatory
actions by HEW prescribing safe conditions of use for food additives. This
legislation defined a food additive as any substance the use of which may
reasonably be expected to result directly or indirectly in its becoming a
component of or otherwise changing the characteristics of any food. However,
the law specifically excluded pesticide chemicals, pesticide residues on
crops (raw agricultural commodities) and pesticide residues in processed food
when such residues result from legal uses of pesticides on crops. Pesticides
continued to be regulated under the 1954 Pesticide Chemicals Amendment,
Although the 1958 Food Additives Amendment intentionally excluded pesti-
cides, it contained a clause on carcinogens which is highly significant for
this study. This provision, known generally as the "Delaney Clause", stipu-
lates that no additive shall be deemed safe if it is found to induce cancer
when ingested by man or animals.
The 1958 Food Additives Amendment represents the first legislation deal-
ing directly with carcinogens in food products. It does not permit any con-
sideration of the level of risk involved, nor does it permit any considera-
tion of potential benefits from use of the additive in question. The FDA
must ban any additive found to be the least bit carcinogenic. While the
Delaney Clause may not be representative of the stance that all legislative
bodies will assume with regard to such problems, it is an indication that
when elected bodies directly confront such issues, they tend to opt for de-
cisions involving minimum risk, regardless of other potential benefits or
consequences.
The Delaney Clause is not part of any legislation that pertains directly
to pesticides. However, standards used in adversary proceedings involving
pesticides appear to be gravitating towards this position. (See Chapter 7
for further discussion of this development.)
47
-------
Current legislation concerning pesticides. TJ During the 1960's the
environmental movement gathered momentum and became widespread throughout the
United States. By 1970, public concern with environmental matters resulted
in the establishment of the Environmental Protection Agency (EPA) . 8_/ In
1972, Congress enacted major amendments to the FIFRA of 1947 through the Fed-
eral Environmental Pesticide Control Act (FEPCA) . _9_/ These amendments made
EPA (instead of the Department of Agriculture) responsible for administering
FIFRA, and it required that all pesticides sold in the United States — not
just those entering interstate commerce — be federally registered. Regis-
tration under standards of FEPCA is required for all pesticides already in
use as well as new ones.
Reregistration of existing pesticides with EPA was to be completed by
October, 1977. Under this act pesticides are to be classified for either
general or restricted use. For the first time, pesticide users are made
responsible for following the approved direction — that is, applicators who
fail to follow directions, or who use a pesticide for an unauthorized purpose,
are guilty of a misdemeanor. Prior to 1972 "authorized" uses were essentially
advisory in nature, and unauthorized uses were not prohibited in any meaning-
ful sense. In short, FEPCA transformed the earlier FIFRA from a labelling
law into a comprehensive regulatory statute.
Most significantly for the purpose of studying public approaches to low-
probability, high-consequence outcomes, FEPCA gave the Administrator of EPA.
the authority to reject an application for registration altogether, or to
restrict, suspend, or cancel the registration of any pesticide that may have
"unreasonable adverse effects on the environment." This is defined under the
law to mean "any unreasonable risk to man or the environment, taking into
account the economic, social, and environmental costs and benefits of the use
of any pesticide" (emphasis added). 10/ On first reading, this legislative
standard appears less restrictive than the Delaney Clause in the Food, Drug
and Cosmetic Act, since it calls for — or at least permits — a weighing of
all risks, costs and benefits, and since considerable evidence is entered
into the record during a cancellation or suspension hearing. Yet EPA regula-
tions and court precedents both appear to be moving toward the position that
finding a pesticide to be carcinogenic to animals or man constitutes prima
facie evidence of an unreasonable adverse effect.
In 1975 the basic FEPCA legislation was amended 11/ to provide for minor
changes, which made the decision process of EPA more open to outside inspec-
tion and consultation. These amendments require the EPA Administrator to
give notice to and consult with the Secretary of Agriculture on the cancella-
tion or reclassification of any pesticide prior to publication of notices of
intent. Further, Congress specifically required EPA to consider the impact
of such actions on "production and prices of agricultural commodities, retail
food prices, and otherwise on the agricultural economy." 12/ Likewise, regu-
lations concerning administrative procedures before EPA, in both their pro-
posed and final forms, require notice to and consultation with the Secretary
of Agriculture.
The 1975 amendments also created a Scientific Advisory Panel to "comment
as to the impact on health and the environment" of proposed specific actions
48
-------
and general regulations. Notice to and consultation with this panel is man-
dated on the same terms as those specified for the Secretary of Agricul-
ture. 13/
The regulatory decision process created by FEPCA involves a number of
stages which, for purposes of discussion, we will refer to in this study as a
sequential decision making process. Briefly, the various stages are as
follows.
1. Registration or reregistration. This is the initial
decision applying to pesticides not already registered
and to all pesticides registered prior to the enactment
of FEPCA in 1972. It includes the following clusters
of preliminary or subsidiary decisions:
a. Issuance of regulations governing the type and amount
of data required to be submitted in order to register
or re-register any pesticide. Applicants for regis-
tration must submit any information requested by the
agency.
b. For individual pesticides, processing and considera-
tion of the data and information submitted in res-
ponse to the registration or re-registration re-
quirements, leading up to the key decision whether
to register or re-register the pesticide, whether to
do so for general or restricted use, and what, if
any, restrictions to impose.
c. Prior to the EPA Administrator's final decision on
registration and classification, the Agency issues
either a "Rebuttable Presumption Against Registra-
tion" (RPAR), in the case of suspect products, or
an intent to register, in the case of apparently
acceptable products. Publication of these notices
in the Federal Register provides both registrants
and interested third parties with an opportunity to
submit materials bearing on each case.
2. Monitoring. This involves a series of administrative
decisions within the Environmental Protection Adminis-
tration, relating to the organization of its programs
and coordination with monitoring programs implemented
by other agencies. The decisions concern information
gathering, organization, and evaluation for pesticides
already registered and in use. These decisions under-
lie all further decisions initiated by the Agency in-
volving outside parties.
3. Issuance of a Notice of Intent to Cancel registration.
This decision results from the above monitoring pro-
cess, either under the Agency's own initiative, or in
49
-------
response to the intervention of third parties
acting on behalf of the alleged public interest.
Unless the parties concerned acquiesce in the
action, this decision triggers administrative law
proceedings. A closely related variant provided for
by the law is a decision by the Administrator to call
a hearing to assist him in deciding on the cancella-
tion question, without issuing a notice of intent to
cancel. This provides a forum for the submission and
detailed clarification of evidence without placing as
much of a stigma on a product as a formal notice of
intent to cancel.
4. Issuance of a Notice of Intent to Suspend registration.
This decision involves considerations similar to those
noted under (3) above, and is designed for situations
of "imminent hazard" pending the outcome of formal
cancellation proceedings. A notice of intent to sus-
pend must be preceded or accompanied by issuance of
a formal notice of intent to cancel.
5. Suspension of registration. This decision applies to
some or all uses of a pesticide. Suspension is a
temporary measure, pending the outcome of formal
cancellation proceedings. Suspension of a registration
may be preceded by an "expedited" administrative law
proceeding, or the Administrator may, at his option,
suspend registration without any formal hearing.
6. Cancellation of registration. This decision perman-
ently prohibits some or all of the uses for which a
pesticide has been registered. An administrative law
proceeding and review by the Administrator of the
Agency are required, if requested by an affected party.
Beyond step 6 above, parties who wish to take exception to an adminis-
trative decision may appeal to the federal courts.
Stages 1 and 2 were described in the introduction to this chapter as
"regulation," and stages 3 through 6 collectively fall under "adversary pro-
ceedings". The following sections of this chapter describe in more detail
how each stage functions. The analysis of theoretical issues concerning this
sequential decision process is postponed until Chapter 8.
Administrative Regulation of Pesticides^
Under the 1972 FEPCA legislation, all pesticides used in the past had to
be reregistered by October, 1977, in order to be legally marketed after that
date. Rules and criteria governing the registration, reregistration, and
classification procedures for pesticides were published in the Federal Reg-
50
-------
ister on July 3, 1975 (pp. 28242-28286). Registration and reregistration of
individual pesticides is also underway as an ongoing activity.
Of the 35,000 pesticides used in the past, some are no longer in use or
are no longer commercially important. Nevertheless, EPA's reregistration
workload has been staggering. In order to facilitate processing such a
large number of applications, the Office of Pesticide Programs (OPP) within
EPA reviews the initial information that registrants submit (which now in-
cludes data and results of animal tests pertaining to carcinogenicity), and
then publishes in the Federal Register either (1) a Notice of Intent to Regis-
ter, in the case of products which it finds acceptable; or (2) a Rebuttal Pre-
sumption Against Registration (RPAR), in the case of suspect products. Any
interested party wishing to reverse EPA's tentative decision has an opportun-
ity at this point to submit materials bearing on the case. This completes
the first phase. However, as noted by Phillip L. Spector in a study of EPA
pesticide regulation: 14/
Under the statute, registration of a pesticide does
not end the inquiry concerning its environmental safe-
ty. The registrant is under a continuing statutory
duty to submit any information it obtains concerning
unreasonable adverse effects. Other interested persons
may provide EPA with relevant information, and the
Agency's own scientists frequently review the available
literature and conduct research on registered products
and their chemical constituents. Hence there are sev-
eral sources from which the Administrator may obtain
new information that changes his previous assessment,
made at the time of registration, that the benefits of
a pesticide's use outweighed its risks. Moreover, a
change in Administrators or in societal values may lead
to a fresh evaluation of the seriousness that should be
ascribed to predicted risks.
Monitoring program. Follwing the initial cluser of decisions involving
registration and reregistration, the second stage moves on to monitoring
activities. This includes information-gathering activities by EPA, such as
reading and reviewing scientific articles and reports on pesticide experi-
ments, especially those concerned with environmental impacts. Of greater
importance, however, is a multi-agency pesticide monitoring program in which
EPA plays a central role. While several government agencies conducted cer-
tain monitoring studies prior to 1972, the FEPCA legislation mandated the
Administrator of EPA to formulate a national monitoring plan: 15/
Section 20(b) NATIONAL MONITORING PLAN: The Adminis-
trator shall formulate and periodically revise, in coop-
eration with other Federal, State, or local agencies,
a national plan for monitoring pesticides.
Section 20(c) MONITORING: The Administrator shall under-
take such monitoring activities, including but not limited
to monitoring in air, soil, water, man, plants, and animals,
51
-------
as may be necessary for the implementation of this Act
and of the national pesticide monitoring plan. Such
activities shall be carried out in cooperation with other
Federal, State and local agencies.
EPA is now developing a comprehensive plan for monitoring pesticide resi-
dues in both human and animal tissue, as well as in the soil, water and air.
While EPA will coordinate the plan and act as a central clearinghouse for the
information gathered, the plan will involve several other government agencies
on a cooperative basis, such as the Department of Agriculture (soils and
crops; meat and poultry); the Department of Health, Education and Welfare
(residues in food and feed; dietary intake; surveillance of crops and commod-
ities) and the Department of Interior (water and wildlife). 16/
With the national pesticide monitoring plan well underway, EPA's own
monitoring activities are increasingly involved with the daily mechanics of
the regulatory process. In the past, these monitoring activities have been
oriented toward the collection and dissemination of data in a widely useful
published form, but the regulatory process is generating pressures for
special-purpose data serving immediate internal needs of the Agency, at the
cost of de-emphasizing publication. Since EPA's budget and resources are
limited, it may be forced to trade-off between more timely information which
is prepared only in "rough" form for extremely important internal uses versus
more delayed information which is published in a more finished format. This
is happening precisely at a time whan the Freedom of Information Act is ex-
posing all monitoring functions to a rising crossfire of outside demands for
data. To cope with such conflicting demands, much tighter coordination of
all internal functions may be required.
In response to such pressures, an internal reorganization and coordina-
tion process is now underway within EPA (see below). As a part of this pro-
cess, EPA's Office of Pesticide Programs is designing and implementing a
Hazard Evaluation System (HES) Program to complement other monitoring efforts.
This program (discussed at greater length in Chapter 8) is concerned with the
collection, verification and analysis of data, as well as with alerting and
triggering mechanisms to be activated when the data indicate that special or
previously unforeseen hazards are developing.
Adversary Proceedings Conducted by EPA
Since enactment of the 1972 FEPCA legislation, EPA has initiated cancel-
lation actions against a number of pesticides. In several important cases,
including DDT, aldrin/dieldrin and chlordane/heptachlor, an important issue
— or even the central issue — was the question of whether residues of the
pesticide or their metabolites are carcinogenic. The case of chlordane/
heptachlor is discussed in more detail in Chapter 7.
As indicated previously, notices of cancellation and suspension are
issued by the Administrator of EPA. The Administrator receives information
and suggestions for initiating such actions from both within and outside EPA.
In 1970, shortly after EPA was formed, the Environmental Defense Fund (EOF)
52
-------
petitioned the Administrator to cancel aldrin/dieldrin, and EDF has been a
major outside influence in initiating action against several other pesti-
cides, including DDT. Within EPA, the Office of Pesticide Programs (OPP) has
the day-to-day responsibility for evaluating registration data, monitoring
results and advising the Administrator if a hearing concerning cancellation
of a pesticide's registration appears warranted. After a formal Notice to
Cancel has been published in the Federal Register, affected parties have 30
days in which they may, as a matter of right, request a formal hearing. When
a hearing is requested, preparation and presentation of the case is handled
by the legal staff members in the Office of General Counsel (OGC) who act as
proponent of the Agency's position and treat OPP as its client. 18/
In some instances, following issuance of a cancellation notice, no re-
quest for a hearing was received and cancellation then became "final and
effective." This occurred, for instance, with some of the mercurial and
arsenical pesticides. In the case of widely used pesticides such as DDT or
aldrin/dieldrin, however, a cancellation notice will almost inevitably result
in a petition for a formal hearing, and the elapsed time from issuance of the
cancellation notice to the Administrator's final order may run many months,
frequently years. Unless a suspension notice is issued, manufacture and use
of the pesticide may continue at least until the Administrator issues a final
decision, and beyond that time if the Administrator's decision is appealed in
the courts.
A difficult problem is faced by an agency which must regulate low-
probability risks whose lifetime incidence may vary between, say 10 and
10 per person _19/ and where direct cause and effect are not and can not be
known with certainty. It should be noted that EPA now uses different cri-
teria for the various decisions which it must make, such as initiating an
action or issuing a final administrative order. In general, the courts have
sanctioned this approach. In the early stages, for example, if the Adminis-
trator has reasonable suspicion for questioning whether a pesticide is car-
cinogenic, the courts have held that under the law he has a responsibility to
issue a Notice of Intent to Cancel the registration, thereby triggering the
creation of a public forum for the issue. Specifically, the courts have held:
The legislative history supports the conclusion that
Congress intended any substantial question of safety to
trigger the issuance of cancellation notices .... 20/
Public hearings bring the public into the decision-making
process and create a record that facilitates judicial
review. If hearings are held only after the Secretary
is convinced beyond a doubt that cancellation is neces-
sary, then they will be held too seldom and too late in
the process to serve either of these functions effect-
ively. 21/
. ... if the Administrator has a substantial doubt as
to safety, it is his duty ... to issue the cancellation
order. (emphasis added) 22/
53
-------
This means that in the face of conflicting scientific evidence about
possible carcinogenicity, the Administrator has a responsibility to initiate
action to cancel the registration of the pesticide in question. Neither
definitive evidence nor a weighing of risks versus benefits is required at
this point.
A decision to issue a Notice of Intent to Suspend registration must be
based on suspicion plus reason to believe that an immediate hazard will
result from continued use of the pesticide while the cancellation proceeding
is in progress. The Administrator need not weigh risks versus benefits when
deciding whether to issue a Notice of Intent to Suspend. However, a final
decision to suspend can stand only if there is no evidence to substantiate
that the original suspicion was unfounded, with the burden of proof placed on
the registrant. A weighing of risks versus benefits must underlie any deci-
sion to suspend.
A final decision to cancel registration permanently results from a con-
sideration of the most solid evidence of all, since the cancellation pro-
ceeding allows evidence to be developed in more detail. Burden of proof that
the pesticide is safe rests with the registrant at all times, but the Adminis-
trator is obligated to weigh benefits against risk when making a final deci-
sion. This decision process will be analyzed in Chapter 8.
As a result of the record established during prior hearings and court
appeals, we now have extensively documented public records of how adminis-
trative law judges and courts deal with issues raised by the possible carcin-
ogenicity of pesticide residues. A less documented link in the total regu-
latory process, however, is the series of events which must necessarily
precede the formal action that initiates an adversary process (i.e., issuance
by the Administrator of a Notice of Intent to Cancel or Suspend the pesti-
cide's registration). This process, which is still evolving within EPA, is
also less open to public inspection than either administrative law proceed-
ings or court records. Therefore, this case study has devoted particular
attention to the following questions:
* What groups within EPA receive information concern-
ing possible hazardous effect of pesticides?
* Who within EPA is responsible for analyzing informa-
tion on hazards and initiating action to recommend
cancellation or suspension of a pesticide where the
information indicates that such action may be desirable?
* How much other information do those responsible EPA
units receive? What are their other duties and re-
sponsibilities? Is it possible or likely that a flood
of information to be processed, plus other organiza-
tional duties and responsibilities will prevent mean-
ingful attention to pesticides that constitute potential
problems?
54
-------
* From a public policy viewpoint, to what extent should
outside organizations (such as EDF) be relied upon to
act as key catalysts in initiating action on behalf
of the public interest? Should such agencies receive
some kind of public funding, such as attorneys' fees,
for actions which are ultimately successful?
* Can Congressional and political pressures be expected
to influence EPA toward excessive or insufficient
caution? Should EPA be given more or less indepen-
dence? Are there organizational modifications that
would promote a better average outcome?
Notes to Chapter 6
1. U.S.C. 1964 Title 7, Sect. 135 et_ se%. April 26, 1910, c. 191, 36 Stat.
331.
2. U.S.C. 1964 Title 21, Sects. 301-392, June 25, 1938, c. 675, 52 Stat.
1040.
3. U.S.C. 1964 Title 7, Sect. 135 et_. seq. June 25, 1947, c. 125, 61 Stat.
163.
4. U.S.C. 1964 Title 7, Sects. 135-135b, 135f, 135g May 12, 1964, 78 Stat.
190.
5. 83rd Congress, Public Law 518, c. 559, 511, July 22, 1954.
6. 85th Congress, Public Law 929, 72 Stat., 1874, Sept. 6, 1958.
7. A good survey of this legislation from the legal point of view is found
in Phillip L. Spector, "Regulation of Pesticides by the Environmental
Protection Agency" Ecology Law Quarterly (1976), pp. 233-8.
8. Spector, "Regulation of Pesticides. . ." p. 233 fn. 2.
9. 92nd Congress, Public Law 516. See 7 U.S.C. Sect. 135 et sec^. (Supp.
Ill, 1973).
10. Spector, "Regulation of Pesticides. . ." p. 235 fn. 13 comments:
This statutory standard is so vague as to amount to essen-
tially no standard at all. At one time it might have been
considered an unconstitutional delegation of legislative
power, although it would probably be constitutional today. .
As one response to the problem of broad delegation,
courts have required agencies to limit their own discretion
by adopting more precise rules. . . EPA recently promul-
gated regulations that give somewhat more content to FEPCA's
'unreasonable adverse effects' test.
55
-------
11. 94th Congress, Public Law 140, 89 Stat. 751, 7 U.S.C.S. Sects. 136-136b,
136p, 136s, 136w, 136y, November 28, 1975.
12. 94th Congress, Public Law 140, 89 Stat. 751, 7 U.S.C.S. Sects. 136-136b,
136b, 136p, 136s, 136w, 136y, November 28, 1975.
13. 94th Congress, Public Law 140, 89 Stat. 753.
14. Specter, "Regulation of Pesticides. . ."p. 236.
15. 92nd Congress, Public Law 516, 86 Stat. 973.
16. A detailed listing of Federal pesticide monitoring activities and sam-
pling locations is presented in: Federal Working Group on Pest Manage-
ment, Monitoring Panel Catalog of Federal Pesticide Monitoring Activi-
ties in Effect July 197_3 (March 1975). See also:
Spencer, Donald A. The National Pesticide Monitoring
Program, an Overview of the First Ten Years of the Pro-
gram's Operation (Washington, D.C.,: National Agricultural
Chemicals Assn., 1974). This work includes a foreward by
G. Bruce Wiersma, Branch Chief, Ecological Monitoring
Branch, EPA.
U.S. Environmental Protection Agency, Office of Pesticide
Programs, "The National Pesticide Monitoring Plan — Draft"
September 11, 1975. This was prepared in response to the
1972 FEPCA legislative mandate and was seen only in draft
form.
U.S. Environmental Protection Agency, Office of Pesticide
Programs, Ecological Monitoring Branch, Pesticide Moni-
toring, a Quarterly Report. This serial began with a
combined issue (No. 1-2) for July/December 1975.
17. Under the proposed reorganization there will be an entire Hazard Evalu-
ation Division to implement the HES Program. EPA has commissioned sev-
eral outside studies on related projects, some of which were completed
in time for our consideration. The most directly relevant report in
this group is:
U.S. Environmental Protection Agency, Office of Pesticide
Programs, A Study to Develop Alternative Conceptual Approaches
for a Pesticide Hazard Evaluation System (Cambridge, Mass.:
Arthur D. Little, Inc., April 1976).
Other studies are:
U.S. Environmental Protection Agency, Office of Pesticide Pro-
grams, A Benefit-Cost System for Chemical Pesticides (Washing-
ington, D.C.: Environmental Protection Agency, June 1975).
56
-------
U.S. Environmental Protection Agency, Office of Pesticide
Programs, Alternative Futures for Environmental Policy
Planning: 1975-2000 (Washington, D.C.: Government Print-
ing Office, October 1975).
18. Spector, "Regulation of Pesticides. . ."pp. 238-9. We con-
firmed this point in our interviews with EPA officials.
19. The meaning of these incidence levels can be illustrated as follows.
"If our experiments are conducted in animals for their whole lifetimes
(as they often are) and if we assume that results on animals are directly
convertible to man (which is reasonable in the absence of data to the
contrary), then the dose we set for a virtual safety level of 1 in
Q
10 should give us a maximum of 2 additional cancers in a lifetime of
individuals in a population of a size roughly that of the present United
States." (Cited from Nathan Mantel and Marvin A. Schneiderman, "Esti-
mating Safe Levels, a Hazardous Undertaking" Cancer Research (June 1975
Q Q
p. 1380.) 1 in 10 is an incidence of probability of 10 per person.
-4
At a level of 10 , 20,000 additional cancers would result in a U.S.-
size population while at a level of 10 , there would be a 2% lifetime
risk for a single cancer in the entire population of the same size.
20. Environmental Defense Fund, Inc., v. Ruckelshaus, 439 F. 2d 584, 593-4
(1971).
21. Environmental Defense Fund, Inc., v. Ruckelshaus, 439 F. 2d 584, 595
(1971).
22. Dow Chemical Company v. Ruckelshaus, 477 F. 2d 1317 (1973).
57
-------
7. Issues Raised in the Chlordane/Heptachlor
Suspension Case
Chlordane has been manufactured and sold commercially since 1947, and
heptachlor has been in use since about 1950. Under the 1947 Federal Insecti-
cide, Fungicide and Rodenticide Act, which was in effect when these two pesti-
cides were introduced, the Department of Agriculture did not have authority
to deny or restrict the registration of any pesticide. For this reason
chlordane and heptachlor were in use for many years before their effect on
the environment was called into question.
The Widespread Use of Chlordane and Heptachlpr
Technically, chlordane and heptachlor are both rather complicated insec-
ticidal mixtures of chlorinated hydrocarbons. They persist in the soil for
up to three years, and have been found to be effective against a number of
pests capable of causing extensive damage to agricultural crops, damage to
homes (termites), or irritation to humans (fire ants).
Usage of chlordane and heptachlor grew steadily from their introduction
until the end of 1975, when their registration was partially suspended by the
Administrator of EPA. I/ By 1975, chlordane had become one of the leading
insecticides in the United States, and accounted for almost six percent of
total insecticide use. The record of the suspension hearing and the Adminis-
trative Law Judge's opinion contain a lengthy and detailed description of the
specific pests, crops, and locales for which chlordane and heptachlor are
effective. 2J Table 2 indicates the number of pounds used annually, along
with an approximate breakdown of the major types of usage. The principal
agricultural use of these two insecticides is on corn; together they are used
on about 3.5 percent of the total corn acreage in the United States. Prior
to their suspension, over 23 million pounds of these two chemicals were being
discharged into the environment each year.
In home and garden usage chlordane is more important than the data in
Table 2 might indicate. In 1972, chlordane accounted for almost 23 percent
of all insecticides used in homes and gardens. It also represented 30 percent
of all industrial/commercial insecticides used at such places as factories,
plants and airports.
Registration of the closely related pesticides aldrin and dieldrin was
suspended in 1974, _3_/ and the suspension affirmed by the Circuit Court of
Appeals on April 4, 1975. The suspension of aldrin and dieldrin no doubt
59
-------
increased the use of chlordane and heptachlor in those applications where
they were considered acceptable substitutes.
TABLE 2
Uses of Chlordane and Heptachlor, 1973 - 1974
(millions of pounds)
Chlordane Heptachlor
Use
1973
1974
1973 1974
Corn
Potatoes
Other vegetables
Home lawn and garden
Turf
Seed dressing
Strawberries
Tomatoes
Ornamentals
Percent classified otherwise
1.94
0.98
0.06
5.77
1.14
—
0.15
0.30
0.23
7.41
4.32
1.11
0.07
6.33
1.25
—
0.17
0.33
0.25
7.34
1.13 1.19
—
—
—
—
0.19 0.27
—
—
—
0.61 0.55
Fire ants, misc. — — 0.03 0.04
Total 17.98 21.18 1.97 2.05
Source: Recommended decision of the Administrative Law Judge in the hearing
to suspend the registration of chlordane and heptachlor, December 12,
1975, EPA, FIFRA Doc. 384, pp. 20 & 22.
The Residue Problem
Under the mandate of the 1972 FEPCA legislation, as indicated in Chapter
6, EPA has been formulating a National Pesticide Monitoring Plan to coordin-
ate existing monitoring programs of other agencies with those activities
undertaken directly by EPA.
The National Pesticide Monitoring Plan [NPMP] calls
for development of a capability to integrate Regional
60
-------
and State networks with Federal Networks. . . These
systems are expected eventually to enter data into the
National Environmental Pesticide Data System (NEPDS).
. . Current NPMP networks [were] originally designed
and activated in response to the President's Science
Advisory Committee Report in 1963. They are coordinated
through the Monitoring Panel, FWGPM [Federal Working
Group on Pest Management].
Thus, while Section 20 (b) and (c) of the amended FIFRA
has given the first legislative mandate to develop and
operate a monitoring plan and system, such a system has,
in fact, been operational for several years. The existing
NPMP networks, with possible modifications, will be con-
sidered integral parts of the Plan. Their utilization
represents the most efficient and effective alternative
available in devising the Plan.
Both EPA and the Monitoring Panel are reviewing existing
networks and modifying [them] when necessary, kj
The Monitoring Panel, FWGPM, operates under executive
order through the Council on Environmental Quality, and
supports the existing NPMP Networks in an advisory capa-
city. With the enactment of FIFRA, as amended, broadened
operational coordination responsibilities are delegated
to the Administrator of the Environmental Protection
Agency. The Plan, therefore, calls for creation of an
interagency Pesticide Monitoring Coordination Council
(PMCC) to coordinate operational aspects of all pesticide
monitoring activities conducted in fulfillment of Section
20 (b) and (c). PMCC would be comprised of those indivi-
duals directly in charge of the monitoring program of the
involved agencies. Council members would necessarily be
authorized to represent their agency in matters of coordi-
nation and be provided adequate resources from their res-
pective agencies to do so. As mandated, EPA would take the
lead in this cooperative effort. Understandably, such
cooperation could result in the need for formal inter-
agency agreements. _5/
Findings from the various existing monitoring programs and special moni-
toring studies indicate that chlordane, heptachlor, and their derivatives and
metabolites have an ubiquitous presence in humans throughout the United
States. Highlights of these results and analyses, summarized here, indicate
the nature and extent of the problem.
Human Adipose Tissue. The National Human Monitoring Program found posi-
tive amounts of oxychlordane in at least 92 percent of the adipose tissues
analyzed between 1971 (the first year the survey had the capability to detect
oxychlordane residue in human adipose tissue) and 1974. 6_/ Results of this
survey are summarized in Table 3. The concentration of oxychlordane in human
61
-------
adipose tissue has been found to increase with age, an additional finding not
reflected in Table 3.
TABLE 3.
Oxychlordane Residue Found in Human Adipose Tissue, 1971 - 1974
Year
1971
1972
1973
1974
Percent
Positive
93.27
92.29
98.35
98.55
Percent
Trace
0.95
2.71
1.01
1.22
Source: Recommended decision of the Administrative Law Judge
in the hearing to suspend the registration of chlor-
dane and heptachlor, December 12, 1975, EPA, FIFRA
Doc. 384, p. 10
During the years 1970 - 1974, heptachlor epoxide was found in over 90
percent of the human adipose tissues analyzed in the National Human Monitoring
Program. Results of this program are summarized in Table 4. As with oxy-
chlordane, the concentration of heptachlor epoxide increased with the age of
the subject in every year of the survey with the exception of the data
collected in 1974.
TABLE 4
Heptachlor Epoxide Residue Found in Human Adipose Tissue, 1970 - 1974
Year
1970
1971
1972
1973
1974
Percent
Positive
94.76
96.23
90.26
97.71
96.33
Percent
Trace
3.61
3.47
2.86
1.92
3.45
Mean
ppm
0.17
0.12
0.12
0.12
0.10
Maximum
PJDTO
10.62
1.53
1.21
0.84
0.77
Source: Recommended decision of the Administrative Law Judge in
the hearing to suspend the registration of chlordane and
heptachlor, December 12, 1975, EPA, FIFRA Doc. 384, pp. 9-10.
62
-------
Human milk. An analysis of 57 human breast milk samples taken in 1974
found quantifiable levels of heptachlor epoxide in 35.1 percent of all sam-
ples. Quantifiable levels of oxychlordane were found in 45.6 percent of all
samples. One hundred percent of the milk analyzed contained at least trace
amounts of either heptachlor epoxide or oxychlordane (a trace represents a
positive finding of residue which cannot be quantified). 7J
Human fetuses. Heptachlor epoxide was also found to cross the placental
barrier and enter the human fetus. A study of residues of various chlorinated
hydrocarbons in different tissues from stillborn infants found measurable con-
centrations of heptachlor epoxide in adipose tissue, spinal cord, adrenals,
lungs, heart, liver, kidney, and spleen. Heptachlcr epoxide was identified
in the cord blood of 27 out of 30 normal newborn infants. 8/
Urban soil. The Urban Soils Monitoring Program detected residues of
chlordane in the soils of all 37 cities sampled between fiscal years 1970 and
1974; heptachlor epoxide was detected in 28 cities and heptachlor in 13
cities. In many cities, chlordane was detected in approximately 20 to 40
percent of all samples taken, and in several cities heptachlor epoxide resi-
dues were present in 10 to 20 percent of the soil samples. The data from the
Urban Soils Monitoring Program thus indicate that residues of chlordane, hep-
tachlor and heptachlor epoxide are widespread in urban areas, reflecting the
common use of chlordane in and around human dwellings for control of lawn and
garden pests. _9_/
Water and shellfish. The widespread usage of chlordane and heptachlor
for gardens and turf creates a certain amount of runoff which introduces resi-
dues into streams, rivers, lakes, and estuaries. A study of the Chester River
in Maryland established that the alpha and gamma isomers of chlordane were
present in its shellfish, finfish, crabs and sediment. The average concentra-
tions found in this study were relatively small. However, laboratory experi-
ments have demonstrated that shellfish readily accumulate chlordane to high
levels from very low concentrations in the water. For example, when exposed
to a level of about 1 part per trillion in the water, oysters concentrated
chlordane 160,000 times; under similar circumstances, soft-shelled crabs con-
centrated chlordane 35,000 times. 10/
Everyday foods. Finally, the National Pesticide Monitoring Program on
Foods has found that residues of heptachlor and heptachlor epoxide appear
most frequently in foods of animal origin; that is, in dairy products, meat,
fish and poultry. Based on the total diet monitoring program, the percent
incidence of heptachlor-heptachlor epoxide in all food classes has remained
relatively unchanged for the 1965 - 1974 period, but the calculated daily in-
take in mg/kg body weight has gradually but steadily declined during the same
period. The food residue levels of heptachlor, heptachlor epoxide and chlor-
dane are low. Heptachlor-heptachlor epoxide residues were present in 0.8 to
2.0 percent of the average daily intake for 1970 - 1974. Chlordane residues
comprise less than 1/100 of 1 percent of the ADI. Chlordane findings in the
surveillance program, although still low, have increased recently; a change
perhaps due to improved analytical techniques, but more likely to the recent
increase in the agricultural use of chlordane. ll/
63
-------
Evaluating the. Risk of Cancer
The record of the chlordane/heptachlor suspension hearing provides a good
illustration of the difficulties in evaluating the risk of cancer and the
hazard to society posed by such risks. One of the more significant problems
is that with a disease like cancer, any thought of carrying out controlled
experiments on humans is abhorrent and it is not feasible to wait for or con-
template "happenstance" epidemiological studies (e.g., the accidental over-
exposure to some group, such as workers in a plant manufacturing or fabri-
cating the chemical). Even if such accidental overexposure were to occur,
epidemiological studies would require decades because of (1) the long gesta-
tion period of cancer (even then the results might be inconclusive because of
the small number of workers involved); (2) the lack of any precise knowledge
about prior or subsequent exposure of the workers to other carcinogens; (3)
the known wide susceptibility of humans to cancer; and (4) our lack of know-
ledge about what causes this susceptibility. In the case of an ubiquitous
pesticide like chlordane or heptachlor, control groups are almost impossible
to obtain. It might be possible to undertake an epidemiological study among
people who live in areas where the use of chlordane and heptachlor are most
concentrated. 12/ However, if we waited until a validated epidemiological
study indicated that the pesticide was definitely carcinogenic, we would be
faced with a serious situation indeed.
Because valid epidemiological studies are not likely, controlled experi-
ments with animals are relied upon for determining the carcinogenicity of a
pesticide. Though such animal experiments are used for a wide range of pur-
poses, extrapolation of the findings of animal studies to man is not straight-
forward because different species of animals react differently to biologically
active substances. JL3_/ In this vein, it is known that humans are more sensi-
tive to certain substances than are the generally used laboratory animals.
The reverse possibility also exists, creating severe difficulties in extrapo-
lating "from an inbred mouse strain to a genetically heterogenous population
such as man." 14/ Among the problems encountered are: variations in doee-
response curves; uncertainty about the proper basis of dose transfer for vari-
ations in body size, e.g., relative weights or relative surface areas; varia-
tions of retention times for materials, based on metabolic rates and ratios
of blood volume to circulation time; variations in plasma protein and tissue
binding; variations in the rate of hepatic metabolism of foreign compounds;
and variations in the generation and repair times of cells, affecting realign-
ment or mutagenic change and the relative stability or variability in the life
environments of the species compared. 157
Still another cause of uncertainty is that even when substances are fed
to animals under controlled circumstances, it is often impossible to state
conclusively that the substance itself is either cancerous or non-cancerous.
The chemical may cause lesions of specific organs — e.g., chlordane and hep-
tachlor affect the liver — and these can cause tumors which in turn may be
benign or malignant. Malignancy is confirmed if metastases are found, but the
absence of metastases does not confirm that a tumor is benign. Moreover,
benign tumors sometimes turn into malignancies; hence benign tumors themselves
pose some risk of cancer, although the risk is not as high as it would be if
the test animals exhibited malignancies. 16/
64
-------
Finally, even if a controlled experiment shows that concentrated doses
of a substance do induce lesions and cancer in animals, there is major un-
certainty as to whether the relationship between dosage and observed conse-
quences is linear. For example, will one-hundredth of the dosage have one-
hundredth the probability of producing a lesion, or is there some threshold
below which minute quantities are inconsequential, and above which risk in-
creases at an increasing rate? Evidence on this issue is almost totally lack-
ing. We do know, however, for some carcinogenic agents, especially ionizing
radiation, that the higher the dosage one is exposed to, and the longer one
is exposed to it, the higher the probability of cancer. Lack of specific
knowledge about possible threshold effects merely increases the uncertainty
about a substance or agent which is possibly carcinogenic. 17/
Based on the varying evidence on carcinogenicity offered during the sus-
pension hearing of chlordane and heptachlor, Chief Administrative Law Judge
Herbert L. Perlman took the following position: 18/
. . .we conclude that heptachlor and its metabolites
appear to be a carcinogen in the mouse. . .and may be
a carcinogen in the CFN rat. . .and that chlordane
appears to be a carcinogen in the mouse. . . [We] feel
very strongly that these matters should be reviewed by
others more qualified with respect to pathology than we
for a more definitive appraisal. In other words we are
hesitantly unwilling at this time to find that heptachlor
and chlordane are conclusively carcinogens in laboratory
animals.
The judge, however, added the following comment: 19/
The Administrator could perhaps do so on the basis of
this record. . . [which] contains no valid epidemio-
logical proof that the pesticides at issue are safe with
respect to the question of cancer and otherwise does not
take into consideration other avenues of exposure to the
chemicals such as inhalation and dermal contact, nor does
it account for the wide variability in human susceptibil-
ity as compared to laboratory animals and different de-
grees of exposure, and considers exposure only to a
simple carcinogen. . . In reality, the attempt to estab-
lish a safe level of a carcinogen for man is in its in-
fancy.
Judge Perlman, in keeping with these reservations, cautioned: 20/
We do not wish to leave the impression that the safety of
heptachlor and chlordane is not still a matter of grave
concern. In other words, they are not "home free" by any
means with respect to whether they are carcinogens in lab-
oratory animals and a potential carcinogenic risk to man.
We merely state that at this time we do not so conclude
65
-------
and, by reason thereof, we cannot find an "imminent
hazard" with respect thereto.
On this basis the Administrative Law Judge recommended that the notice
of intention to suspend chlordane and heptachlor be dismissed and that final
determination of the issues raised by chlordane and heptachlor rest solely on
the outcome of the cancellation proceeding.
The Administrator of EPA did not concur with all of the Administrative
Law Judge's conclusions. His stated reasons for arriving at different con-
clusions were as follows: 2J7
In sum, therefore, the ALJ's precise holding that there
was no "imminent hazard," and thus no basis for suspension
under section 6(c) of the FIFRA (1) was made without balan-
cing the risks of continued use of chlordane and heptachlor
against the benefits of that use, and (2) was grounded solely
on the holding he was "hesitantly unwilling" to find that
chlordane and heptachlor are "conclusively" carcinogens in
laboratory animals even though they both "appear to be" such
carcinogens and the question of their safety on this score is
"a matter of grave concern."
Both of these holdings contain misapplication of the perti-
nent statutory provisions and controlling judicial decisions.
Section 6(c) (1) of the FIFRA provides that the Administra-
tor suspend the registration of pesticides where "that action
is necessary to prevent an imminent hazard during the time
required for cancellation proceedings. . .
The term "imminent hazard" is defined in Section 2(1), in
pertinent part, as "a situation which exists when the continued
use of a pesticide during the time required for cancellation
proceeding would be likely to result in unreasonable adverse
effects on the environment. . .
The phrase "unreasonable adverse effects on the environ-
ment", in turn, is defined in section 2(bb) to mean, "any un-
reasonable risk to man or the environment, taking into account
the economic, social and environmental costs and benefits of
any pesticide."
Thus, in combination, section 6(c)(l), 2(e), and 2(bb)
provide that suspension depends upon a finding that continued
use of a pesticide during the time required for a cancellation
proceeding would be likely to result in any unreasonable risk
to man or the environment, taking into account the economic,
social and environmental costs and benefits of the use of the
pesticide.
66
-------
As the courts have held (and, indeed, as counsel for
the registrant have urged in this proceeding) , the ques-
tion of the '"reasonableness" of the risk presented depends
upon an assessment of the nature and magnitude of risk of
harm to man or the environment as compared to the nature
and magnitude of economic, social, and environmental bene-
fits which are derived from the use of the pesticide
during the time required for a cancellation proceeding. 22 /
Thus, the FIFRA and the cases, in my opinion, make two
points clear with regard to a suspension proceeding. First,
it is not necessary to find "conclusively" that actual harm
to man x^ill occur if the use of the pesticide in question is
continued; rather the finding required is that continued use
during the cancellation proceeding is "likely" to result in
any "unreasonable risk" to man or the environment. As the
court of appeals has held:
[T]he term "imminent hazard" is not limited
to a concept of crisis: It is enough if there
is substantial likelihood that serious harm will
be experienced during the year or two required in
realistic projection of the (cancellation) pro-
cess. 23/
Second, the propriety of suspension turns upon an analysis
in which the risks are balanced against the benefits, rather
than from an analysis of risks or benefits alone.
It should also be noted that the mere fact that the evidence
on either of these issues (risks or benefits) is not complete,
or that more evidence may be expected to be developed in the
cancellation proceeding, is not a reason to deny suspension.
Suspension is an interim remedy, to be determined on a record
assembled on an expedited basis, as was the case here. Con-
gress must have expected that the record of a suspension pro-
ceeding would not contain all the evidence on crucial questions
that would be expected and presented in a cancellation pro-
ceeding. As the court of appeals has held:
[T]he function of the suspension decision is to
make a preliminary assessment of evidence and proba-
bilities, not an ultimate resolution of difficult
issues.
Fcr the foregoing reasons, I find that the ground given by
the Administrative Law Judge in determining not to recommend
suspension of the registration of heptachlor and chlordane is
deficient as a matter of law. He failed to weigh adequately
the risks in relation to the benefits as required by the stat-
ute and by Agency policy. Moreover, he held that the fact that
the carcinogenicity of heptachlor and chlordane had not been
67
-------
"conclusively" established with respect to laboratory
animals was determinative of the potential risk of cancer
to man, even though the statute talks in terms of the like-
lihood of an unreasonable risk. Indeed, I believe that
this much less stringent statutory standard was met by Judge
Perlman's findings that heptachlor and chlordane "appear to
be" carcinogenic in laboratory animals and are pesticides
whose safety is a matter of "grave concern." In the same
vein, the courts have also insisted that the burden of proof
of the safety of a pesticide lies at all times with the
registrant. 25_/ Thus insofar as the showing of risk of harm
from heptachlor and chlordane is not yet conclusive, it
should be emphasized that it is not the Agency's burden to
establish the risk of harm, but rather the registrant's
burden to disprove it once its safety has been called into
question by a sufficient showing of probable risk.
Thus, as shown by the record of this suspension proceeding, the courts
and EPA have adopted a position which lies rather close to the Delaney Clause
used by the FDA in evaluating food additives. That is, the likelihood of a
pesticide causing cancer in animals will be interpreted to imply that it is
also likely to cause cancer in man, and this in general constitutes sufficient
grounds for suspending registration for any use likely to result in residue in
the human food supply.
The key difference between this position and the Delaney Clause is that
benefits have been weighed against risks, whereas the Delaney Clause leaves no
discretion whatever for consideration of benefits. Certain exceptions were
made on the grounds that benefits outweighed risks. The uses most essential
for continued production and whose denial would have concentrated adverse
economic impacts in particular areas — as was found to be the case with the
Hawaiian pineapple crop and a few other minor uses of chlordane and hepta-
chlor — were exempted from the suspension order. Likewise, to prevent im-
mediate, major economic loss to producers, distributors and users, it was
authorized that existing stores of the pesticides could be used during the
1976 crop year.
Some of the theoretical issues raised by this case study will be covered
in the next section.
Notes to Chapter 7
U.S. Environmental Protection Agency, Suspension Proceedings Before the
Administrator, Decision of the Administrator on the Suspension^) f
Heptachlor-Chlordane, in re Velsical Chemical Corporation et al., regis-
trants, FIFRA Docket No. 384, Expedited Hearing on Heptachlor-Chlordane;
dated December 24, 1975, 77pp.
68
-------
2. U.S. Environmental Protection Agency, Before the Administrator, Recom-
mended Decision, by Herbert A. Perlman, Chief Administration Law Judge,
in re Velsical Chemical Corporation et al_._, Registrants, FIFRA Docket
No. 384, Expedited Hearing on Heptachlor-Chlordane; dated December 12,
1975, 122pp.
3. U.S. Environmental Protection Agency, "Opinion and Order of the Adminis-
trator, Environmental Protection Agency, in the Suspension of aldrin-
dieldrin," Federal Register, v. 39, no. 203, Part 1, October 18, 1974,
37265-37272.
4. U.S. Environmental Protection Agency, Office of Pesticide Programs, The
National Pesticide Monitoring Plan, Draft, Sept. 11, 1975, p. 10.
5. U.S. Environmental Protection Agency, Office of Pesticide Programs, The
National Pesticide Monitoring Plan, pp. 12-13.
6. National Human Monitoring Program, cited in U.S. Environmental Protec-
tion Agency, Before the Commissioner, Recommended Decision, p. 10. This
program is incorporated in Environmental Protection Agency, Office of
Pesticide Programs, "National Pesticide Monitoring Plan" pp. 42-43.
7. U.S. Environmental Protection Agency, Before the Commissioner, Recom-
mended Decision, p. 11.
8. U.S. Environmental Protection Agency, Before the Commissioner, Recom-
mended Decision, p. 12.
9. U.S. Environmental Protection Agency, Before the Commissioner, Recom-
mended Decision, pp. 12-13; see also U.S. Environmental Protection
Agency, Office of Pesticide Programs, The National Pesticide Monitoring
Program, p. 17. The Urban Soils Monitoring Program is incorporated in
the EPA plan (see note 4 above) as a component of the National Soils
Monitoring Network which also includes two other components cropland and
noncropland.
10. Cited in U.S. Environmental Protection Agency, Before the Commissioner,
Recommended Decision, pp. 14-17.
11. U.S. Environmental Protection Agency, Before the Commissioner, Recom-
mended Decision, pp. 18-19.
12. The highest use of chlordane occurred in Missouri bottomland, where
farmers applied the pesticide each year as "insurance" against the corn
cutworm. According to the record of the suspension hearing, thirty per-
cent of all corn grown in Missouri, and most of the corn grown on bottom-
land, was treated with chlordane. Locally produced meat, milk, chicken
and eggs might be expected to have higher concentrations of chlordane.
To the extent that those who live on farms and in small rural communities
consume high proportions of locally produced foods, they would have
greater exposure to chlordane.
69
-------
13. Of the several technical articles analyzing carcinogenic risk which have
been introduced into evidence at the suspension hearing, the one which
addresses this issue most directly — and is also by far the most read-
able for policy makers without losing any of its analytic incisiveness —
is "From Mouse to Man — or How to get from the Laboratory to Park Avenue
and 59th Street," by Marvin A. Schneiderman, Nathan Mantel, and Charles
C. Brown, Annals of the New York Academy of Science, (Jan. 31, 1975).
14. Schneiderman, "From Mouse to Man. . ." p. 245.
15. Schneiderman, "From Mouse to Man. . ."p. 245.
16. The transcript of the suspension hearings contains many examples of dis-
agreements by expert witnesses over the nature and significance of
histological and other pathological findings.
17. The fundamental article on the statistical analysis of carcinogenicity
test results is "Safety Testing of Carcinogenic Agents," by Nathan
Mantel and W. Ray Bryan, Journal of the National Cancer Institute
(August 1961) pp. 455-470. The proceedings are reviewed fifteen years
later in "Estimating Safe Levels or Hazardous Underestimating," by
Nathan Mantel and Marvin A. Schneiderman, in Cancer Research (June 1975)
pp. 1379-86. The same ground has been covered exhaustively by the Sub-
committee on Estimation of Risks of Irreversible Delayed Toxicity of the
Committee to Coordinate Toxicology and Related Programs of the U.S.
Department of Health, Education and Welfare. The report of this sub-
committee has been published as "Estimation of Risks of Irreversible
Delayed Toxicity" by David G. Hoel and others in Journal of Toxicology
and Environmental Health (1975) pp. 133-151. The four articles cited
in this note and in note 13 above have been introduced into evidence at
the suspension hearings by the U.S. Environmental Protection Agency.
18. U.S. Environmental Protection Agency, Before the Commissioner, Recom-
mended Decision, p. 86. Emphasis in original.
19. U.S. Environmental Protection Agency, Before the Commissioner, Recom-
mended Decision, p. 86. Emphasis in original.
20. U.S. Environmental Protection Agency, Before the Commissioner, Recom-
mended Decision, p. 87. Emphasis in original.
21. The entire following passage is cited from U.S. Environmental Protection
Agency, Suspension Proceedings Before the Commissioner. Decision of
the Administrator on the Suspension of Heptachlor-Chlordane, pp. 9-12.
22. Environmental Defense Fund v. Environmental Protection Agency, 510 F.
2d 1292, 1302 (C.A.D.C. 1975); Environmental Defense Fund v. Environ-
mental Protection Agency, 465 F. 2d 528, 538 (C.A.D.C. 1972). (footnote
in original).
70
-------
23. Environmental Defense Fund v. Environmental Protection Agency, supra,
510 F 2d at 1297, citing Environmental Defense Fund v. Environmental
Protection Agency^ supra, 465 F 2d at 540. (footnote in original).
?4. Environmental Defense Fund v. Environmental Protection Agency, supra,
510 F 2d at 1297, citing Environmental Defense Fund v. Environmental
Protection Agency, supra, 465 F 2d at 537. (footnote in original).
25. Environmental Defense Fund v. Environmental Protection Agency, supra,
510 F 2d at 1297; Dow Chemical Co. v. Ruckelshaus, 477 F 2d 1317, 1324
(C.A. 8 1973). (footnote in original).
71
-------
8. Theoretical Issues Posed by Potentially
Carcinogenic Pesticides
Products containing chlordane and heptachlor are among the thousands
of pesticide products available through suppliers. The products are bought
for agricultural, garden, and other insect control purposes, by both business
firms and consumers. Farms and other pesticide-using businesses in turn sell
agricultural products, or services such as termite-proofing for housing, to
the ultimate consumer. As a result of this widespread distribution and use
the chemical constituents of chlordane, heptachlor, oxychlordane and hepta-
chlor epoxide now appear almost ubiquitously as a trace component in the en-
vironment, in our food, and in our body tissue. Continued exposure to traces
of pesticide residues may carry low-grade but widespread carcinogenic risks.
One theoretical approach to decision-making under risk is based on the
maximization of expected utility (by households) and expected profits (by
businesses) in the hypothetical framework of a perfectly competitive market.
In this idealized market the forces of supply and demand establish an equi-
librium with a number of desirable properties. Since the attendant risks
with which this study is concerned arise in the context of market trans-
actions between several kinds of sellers and many different buyers, it is
instructive to approach a theory of such risks in a market context. While
the idealization inherent in the neoclassical concept of the market intro-
duces a series of caveats into any such analysis, the theoretical framework
is so deeply entrenched in the thinking of economists that we take it as our
baseline, if for no other reason than to lay bare its limitations.
An alternative theoretical approach to decision-making under risk in-
volves centralized decisions based on a systematic comparison of alternatives.
Hazards to the public are considered in the context of economic, social and
environmental costs and benefits. This approach is embodied in the 1972 Fed-
eral Environmental Pesticide Control Act, which amended the 1947 FIFRA legis-
lation. This legislation requires that the Environmental Protection Agency
employ risk-benefit analysis within the context of what we will refer to as
a "sequential decision-making process."
The purpose of this chapter is to examine the applicability of these
and other possible theoretical approaches to decision-making under risk.
The chapter deals with the following principal topics, illustrated through-
out with examples drawn from specific considerations raised by chlordane and
heptachlor:
73
-------
* Market-based theories of decision-making
* The existing sequential decision process
* Risk-benefit analysis
* Dynamic considerations
Market-based Theories of Decision-making
As indicated above, the central theoretical approach of neoclassical
economics to decision-making is based on maximization of utility and profits
within an idealized market context. \J In a world of perfect competition,
with no risk or uncertainty, consumers are assumed to maximize their indi-
vidual utility functions, and business firms are assumed to maximize profits.
Consumers and business firms both are taken to have complete information
about all relevant aspects of their market transactions. Under these postu-
lates, market prices provide necessary and sufficient information for con-
sumers and producers to maximize their respective utility and profits. If
all benefits and costs of market transactions accrue to the parties directly
involved in the transaction — that is, if the transactions do not have any
external effects associated with them — individual, decentralized decision-
making not only enables individual consumers and business firms to maximize
their respective objectives; it also leads to a socially efficient distribu-
tion of goods and services, given the initial distribution of assets. Effi-
ciency is based on the criterion of "Pareto optimality," which, in brief,
states that an efficient distribution of goods and services is achieved when
there is no way of making one person better off without making another person
worse off.
In a perfectly competitive model, market prices are assumed to reflect
all social costs and benefits of production, and it is assumed that all
social costs and benefits from a market transaction accrue exclusively to
the buyer and seller. If third parties, who are in no way engaged in the
production or consumption of an item, are adversely affected by the produc-
tion and consumption of an item, other than through a change in market prices,
external diseconomies are said to exist. It has long been recognized that
the existence of external diseconomies can cause market failure; i.e., choices
arrived at in a decentralized market will not be socially optimal. Market
failure occurs because prices no longer reflect true social cost, and con-
sumers and producers thus base their decisions on imperfect information. 2/
The existence of external diseconomies raises the issue of whether some
means can be found to internalize social costs, so that individual decisions
in the market can be restored as a desirable means by which individual and
collective decisions are made.
The discussion of external diseconomies is pertinent to pesticides, be-
cause any carcinogenic microscopic residue can be classified as an external
diseconomy. When pesticides are used on a farm, residues from the pesticide
will almost surely persist on some of the farm products; the more persistent
74
-------
the pesticide, the more likely it is that residues will wind up in human con-
sumption and, subsequently, in human tissue. Moreover, residues from almost
all uses, including lawn and garden applications, are washed into streams or
rivers, thereby contaminating the water supply of those downstream. When
contaminated water enters bays or estuaries, some of these residues may be-
come concentrated in shellfish, such as oysters. Those who drink the water
or eat the shellfish will also ingest microscopic amounts of the carcinogen,
which may cause cancer.
Note that our statement is that such a residue "may cause cancer." We
are not dealing with an external diseconomy that arises in every instance
with perfect certainty; rather, it is one that occurs with a low probability,
but whose consequences are serious for anyone who does in fact develop cancer
from exposure to carcinogenic residues. Thus pesticide residues not only
raise issues traditionally associated with external diseconomies; they also
raise issues that arise from considerations of risk and uncertainty. Of im-
portance from a public policy viewpoint is the fact that almost all citizens
are exposed to some amounts of the residue, and that this exposure may even-
tually cause cancer in a sufficient number of people to be deemed cata-
strophic.
The existence of risk and uncertainty can cause market failure even when
there are no external diseconomies. For a neoclassical market to operate in
the presence of uncertainties, perfectly competitive sub-markets would have
to exist in all relevant insurance contracts affecting commodities traded. _3/
Gaps in the insurance market can cause the efficient allocation of resources
to break down. This form of failure can present a serious problem for
market-based solutions. 4/
Statistical uncertainty can be introduced into the perfectly competitive
model in either of two ways. First, it may be assumed that the probability
distributions of all uncertain variables are known, and that consumers and
producers maximize expected values rather than fixed functions. In any parti-
cular instance, the actual value of an uncertain variable may deviate from
its expected value, but this does not alter the fact that consumers and pro-
ducers, making decentralized decisions guided only by their self-interest,
will arrive at the best choices compatible with their uncertain conditions.
Second, under the assumptions of Bayesian statistics, consumers and
producers are taken to have formed initial perceptions or judgments concern-
ing the distribution functions of all uncertain variables. These a priori
distributions need not correspond to the true distributions (which are un-
known). If people use observed outcomes to correct their a priori percep-
tions and if underlying distributions do not change, then over time the a_
priori distributions will approach the true distributions. The results de-
rived from a model incorporating Bayesian assumptions are often similar to
those derived from a model that incorporates classical assumptions. Namely,
decentralized decision-makers select those alternatives that are as good as
any which could be selected under the circumstances, and are as good — at
the limit of a series of iterations — as the ones that could result from
known probability distributions.
75
-------
One crucial difference between a world with known probability distribu-
tions and a Bayesian world, however, is that in a Bayesian world the value of
information is substantially enhanced, to the point where it becomes an eco-
nomic commodity that may have a price of its own. The mere pooling of infor-
mation will have value and the production of certain types of information is
likely to occur under conditions of high fixed and low marginal cost. Indi-
visibilities introduced by fixed costs will in turn tend to undermine decen-
tralized market equilibrium. Another problem can arise if various agents in
the economy have differing perceptions of the probability distributions float-
ing around the world. Should this occur, individual actions may interact and
affect each other in such a way as to leave everyone feeling worse off than
they would feel under a collective solution.
Despite these theoretical considerations, in the eyes of most economists
the introduction of risk and uncertainty does not, per se, demolish the ap-
plicability or intellectual appeal of the market model. The existence of
uncertainty merely implies that outcomes will vary even under similar condi-
tions, and that realized outcomes will not always equal the expected values
of these outcomes. That is, for reasons unknown some people may wind up with
less than both parties expected — e.g., some people will unknowingly buy
"lemons", or products that do not perform up to anyone's expectations, includ-
ing the seller's. When this occurs, actual realized cost and benefits will
typically differ at times by a large amount from expected costs and benefits.
These problems are not new, nor are they restricted to low-probability,
high-consequence occurrences. In order to overcome the very real problems
introduced by risk and uncertainty, and to facilitate the operation of mar-
kets, a number of legal safeguards and risk-spreading institutions have been
created. In many instances these have restored faith in the applicability
and desirability of the market model. We therefore review the potential ap-
plicability of such legal safeguards and risk-spreading institutions to the
risk of cancer posed by pesticides like chlordane and heptachlor.
Propagation of accurate information. Over the years, Congress has en-
acted many laws requiring product labels to contain some minimum amount of
accurate information. The original 1910 Pesticide Act, as well as the 1947
FIFRA legislation, were essentially of this type. Availability of such in-
formation, especially if it is prominently displayed, helps prevent fraud,
deceit and misuse. Admittedly, such laws constrain the actions of producers,
and in so doing appear to violate the freedom of adjustments required by the
theory of a perfectly competitive market. However, this type of legislation
does not displace the market as the central arena of fundamental decisions.
In a world of uncertainty and imperfect information, the principal intent of
such legislation is to help make market decisions more informed. An impor-
tant underlying premise is that potential users, when provided with adequate
information, will act at least as wisely in their own behalf as would anyone
else who made decisions and choices for them. It should be noted, however,
that it is not easy to decide which and how much information is useful and
worthwhile to potential purchasers. Due to decision-making costs, the con-
sumer often finds it more economical to grope for crude indicators of overall
worth to him (e.g., brand name, regular gas, little cars get more miles per
gallon than big cars) than to attempt a more technical comparison of the
76
-------
characteristics of different products. It is often difficult, therefore, for
an expert to decide just what kind of information will be valuable to users
and consumers.
The provision of more and better information has been adopted by Con-
gress in many contexts. For example, this approach has been central to deal-
ing with the health threat posed by cigarettes. The hazard involved in smok-
ing is noted on every pack and has also been widely publicized in the nex^s
media. Individual preferences revealed to date — i.e., continued high con-
sumption of cigarettes — indicate that many people either play down the risk
of cancer or do not impute much utility to increasing the probability of
living longer.
The same general approach of improving information and then permitting
free choices within the context of the market mechanism might also be consi-
dered for pesticides. One possibility would be to label food (such as corn)
according to the amount of potentially dangerous pesticide residues which it
contains, and then permit market prices to adjust accordingly. The price
level of contaminated food would presumably drop so as to reflect the dis-
count which people put on the presence of such residues. If all corn were
sold directly for human consumption this approach would at least be feasible.
It might even be a practicable solution to the residue problem.
Most corn, however, is not directly consumed. If corn containing a
pesticide residue were sold at a lower price, such corn might well be bought
preferentially to feed cattle and poultry. All residue in corn would then
wind up in meat and dairy products, as some of it does now. From a theoreti-
cal viewpoint, one could test all steers, cows, or chickens for residues of
pesticides like chlordane or heptachlor, and provide appropriate labels on
all products along the entire product chain. The cost of any such program,
however, could well exceed the cost of eliminating the offending pesticides
altogether. Consequently, the option of providing consumers with information
that is accurate and sufficient for them to make more informed market choices
does not appear to be practicable. Some other way must be found to deal with
microscopic amounts of pesticide residues that are potentially carcinogenic,
especially those residues which become so widespread in the environment that
they can appear in food products that are totally unrelated to the crop ori-
ginally treated with the pesticide; for example, shellfish, eggs and butter.
Compensation to those injured by uncertain events. In the course of
producing or selling goods and services, events sometimes occur in which a
person suffers ill consequences and economic losses resulting from circum-
stances that were not of his own making. Such events are sometimes freaks
of nature, like violent storms that sink ships at sea; sometimes they are
accidents, such as a train derailment; and sometimes they are caused by the
actions of other people. When a person suffers adverse consequences that
are not socially sanctioned and that can be attributed to someone else's in-
tent or negligence, the law of torts frequently provides a partial remedy in
the form of compensation for losses suffered. That is, the party adversely
affected can sue the party at fault for damages, and defendants who are held
responsible must pay the damages. To the extent that such laws apply, their
effect is to make firms responsible for the consequences of their actions
77
-------
and for the effects of their products, thus giving the firms the motivation
to forestall accidents and lawsuits, especially when potential damages exceed
other gains. 5_/
The law of torts is not the only way in which the social costs of uncer-
tainty and occasional undesirable outcomes are internalized in the production
process. In certain instances where risks must inevitably be incurred even
under prudent management, strict or absolute liability — i.e., liability
regardless of fault — has been imposed on firms in order to help protect and
compensate those who suffer unintended ill consequences. Workmen's compensa-
tion laws, which provide reparations for anyone injured on the job, are of
this type. If a workman is injured on the job, regardless of whether the
accident results from uncontrollable circumstances, or from negligence by
management — or even negligence by the injured worker himself — the law
requires a certain amount of compensation for the person's economic losses.
Most firms carry workmen's compensation insurance to cover their legal obli-
gation, but a few larger firms self-insure. This "no-fault" approach to job-
related accidents not only provides compensation for economic losses but, to
the extent that a firm's rates for workmen's compensation insurance reflect
its accident experience, the system also gives each firm an incentive to con-
duct an aggressive safety and accident prevention program.
Still another legal approach used to shift social costs back to pro-
ducers and compensate consumers is the doctrine of implied warranty, which
is often used to compensate people who have in some way been injured or suf-
fered economic loss from defective products. In many states the legal doc-
trine of strict liability is now applied to cases involving product liability.
Many firms purchase product liability insurance to cover this risk. 6f
In all of these cases — negligence, absolute liability, implied war-
ranty, or strict liability — the law helps shift the social cost of risks
and hazards from the individual consumer who suffers the injury back to the
producer. The market mechanism, with its decentralized decision-making mode,
functions within this legal framework. Regarding pesticides the question is:
Would application or extension of any of these legal doctrines provide an
adequate way of coping with risk of cancer from pesticide residues? The
problem with applying any of these legal doctrines is that even when a person
is known to have developed cancer, it is generally impossible at the current
stage of knowledge, to determine whether the cancer was caused by any of the
many carcinogens in the environment and if so, which carcinogen. Unless and
until it becomes possible to establish such cause-and-effeet relationships,
none of these legal doctrines will lead manufacturers to take account of the
social hazards and costs of producing, selling or using pesticides. Our
ignorance of these relationships is, of course, also a serious handicap in
devising other policy measures.
Direct levies on producers to reflect social costs. One policy approach
suggested for some environmental problems is to impose taxes designed to in-
ternalize the social costs of undesirable environmental effects that arise
from the production process. More specifically, in the case of manufacturing
plants that discharge pollutants into the air or water, the efficacy of an
effluent charge has been widely discussed by economists. This approach sug-
78
-------
gests still another possible way of internalizing the social costs that arise
from the risks of cancer. Namely, instead of issuing regulations or restric-
tions governing various and sundry uses, the government could impose a "resi-
due tax" on the pesticides in question, and then allow market forces to de-
termine the extent to which the pesticide will be used.
The contrast between a residue tax and the present regulatory approach
is interesting. Regulatory decisions, by their very nature, have a dichoto-
mous (on-off) character. In other words, the regulator must decide whether
to register a pesticide — or whether to register it for some uses and not
for others. A residue tax, by contrast, is a continuous variable that can be
set at higher or lower levels, and that can be adjusted upwards or downwards
to take account of emerging information regarding risks. A residue tax
carries no absolute proscriptions , and permits individual decision-makers to
decide, at the margin, whether application of the pesticide is worth the cost.
Under a crude residue tax, the total cost of the additional cancer cases
attributable to a specific residue could be levied on some combination of
pesticide manufacturers, pesticide distributors, and farmers if (1) the
charge were levied on the total residue ultimately entering the human biomass
(which can be estimated); (2) the number of cancer cases resulting from a
unit of residue entering the human biomass could be estimated; and (3) the
charge were levied at the time of injection into the environment rather than
after time lags spread out over many years between pesticide application and
entry of residue into human tissue. This approach is not without difficul-
ties .
In principle, a residue tax should be levied on different uses of a
chemical, since different uses result in varying amounts of residue entering
human tissues. Imposing such a levy poses administrative difficulties be-
cause in a competitive market there is a lack of control by manufacturers
and distributors over many applications; e.g., for lawns, for corn, or for
fire ants. In defense of the residue tax approach, it should be noted that
any attempt at selective regulation also poses a similar set of administra-
tive difficulties. Suppose the Administrator decides to permit some uses and
ban others. Then, since products containing the pesticide will still be
manufactured and sold, there must be some means of policing and enforcing the
rules to ensure that products are used in accordance with the rules.
The information required to impose either a crude or fine-tuned residue
tax is roughly similar and equivalent to the information required for either
crude or fine-tuned regulation. Viewed solely from the perspective of eco-
nomic decision-making, a residue tax would appear to be a viable alternative
to the direct regulation approach mandated by Congress. From a policy-making
viewpoint, a residue tax poses the same type of problem as an effluent charge.
These problems have been discussed at some length by Giandomenico Majone. TJ
To summarize Majone's arguments briefly, he points out that while effluent
charges may appear theoretically attractive, the criteria and models used to
derive or exhibit this attractiveness are typically deficient in their treat-
ment of distributional, political, administrative, and legal issues.
79
-------
The distributional problems raised by a residue tax approach are monu-
mental. Consider, for example, the problem of compensation for damages.
Under our present state of knowledge we are unable to attribute an individual
case of cancer to a specific carcinogen, or to a group of carcinogens. Thus
even if producers, distributors, farmers and other pesticide users could be
made to internalize all costs in proportion to the precise cancer damages
they had originated, the statistical incidence of, say, 5,000 cases per year
could not be attributed to specific individuals whose cancer was caused by
carcinogenic pesticide residues, thus making compensation of particular indi-
viduals impracticable. Hence, even if market prices could be made to accu-
rately reflect all social costs, a serious problem of distribution and equity
would remain.
Market-based theories: summary. Microscopic but widely distributed
pesticide residues that are potentially carcinogenic raise issues associated
with external diseconomies as well as risk and uncertainty. In the absence
of any government action whatsoever, neither market prices nor other infor-
mation available to consumers will reflect the element of risk. Consumers
will thus unwittingly and uncontrollably subject themselves to the risk of
cancer (to an unknown degree) when purchasing and consuming many foods.
Chemical manufacturers, distributors of pesticides, and farmers reap the bene-
fits of increased crop yields and profits made possible by the pesticides,
while passing along to the public the increased risk of cancer, and without
having to take economic account of the suffering and costs involved.
In order to preserve and improve the market as the central mechanism for
arriving at social choices, several widely applicable approaches have been
devised to overcome similar problems: (1) provision of more and better in-
formation to consumers; (2) legal doctrines that provide compensation to
injured parties, such as negligence, absolute liability, implied warranty and
strict liability; and (3) direct levies, such as effluent charges, on the
emission of pollutants into the environment. For a variety of reasons, none
of these represents an attractive policy for internalizing the costs that
arise from low-concentration but widely distributed carcinogens. In the con-
text of an economy that relies principally on a market-based framework of
choice, therefore, the central theoretical problem is to devise appropriate
decision processes that can supplement the market in generating information
and drawing attention to the social costs that result from the risks posed by
potentially carcinogenic residues.
The Existing Sequential Decision Process
The many inherent difficulties of relying on market-based processes to
solve environmental problems were implicitly recognized by Congress when it
enacted the Federal Environmental Pesticide Control Act in 1972. Instead of
attempting to operate through the price system, Congress established a non-
market regulatory mechanism which embodies (1) sequential decision making,
and (2) risk-benefit analysis as the basis for regulatory decisions concern-
ing pesticides.
80
-------
The discussion of theoretical issues in this section is based on the de~
scription of the sequential decision process in Chapter 6. The present section
is limited to the identification of issues arising from this sequential deci-
sion process. The next section will discuss the risk-benefit-approach as it
was applied specifically to the chlordane/heptachlor suspension case, and will
raise theoretical issues associated with risk-benefit analysis involving
pesticides.
Provisions of the current law. A brief recapitulation of the more de-
tailed description given above in Chapter 6 yields the following crucial deci-
sion points within the sequential decision process of EPA:
Step
Number
1.
2.
3a.
Brief Description
Registration, including reregistration of existing
pesticides
Monitoring of pesticides registered and in use
Convening of a formal hearing to assist the Administra-
tor in arriving at a cancellation decision; or
3b. Issuance of a notice of intent to cancel registration
4. Issuance of a notice of intent to suspend registration
5. Suspension of registration
6. Cancellation of registration
For purposes of this analysis, we will treat decision stages 1 through 6
as components of a sequential decision-making process which gathers and ana-
lyzes information in order to improve subsequent decisions. Our general con-
cerns here are with the cost and effectiveness of the process, and the
possibility of improving it.
As was pointed out in Chapter 6, the suspension process is embedded in
the cancellation process, since a Notice of Intent to Cancel must be issued
before or concurrently with a suspension notice. Hence, after a pesticide
has been registered and is already in use, the Administrator of EPA has the
following options: (1) He can issue a Notice of Intent to Cancel, thereby
triggering a formal, detailed, lengthy and expensive process of information
gathering and decision making; (la) As a minor alternative to (1), he can
call hearings to help him decide on the cancellation question, without a for-
mal Notice of Intent to Cancel; or (2) as a major alternative to (1), he can
refrain from formal action and allow the monitoring, gathering and analyzing
of information to continue. In order for the Administrator to issue a Notice
of Intent to Cancel, he must first have developed a perception, based on a
reasonable minimum of scientific evidence, that continued use of a pesticide
constitutes a risk to man or the environment. Thus, stage 2 — monitoring,
gathering and analyzing of information — is a vital link in the chain of
81
-------
decisions. Until recently this stage was the most loosely defined part of
the entire regulatory process; a major effort is now underway to fill this
gap-
Triggering of a formal cancellation proceeding. How does and how should
an administrative bureaucracy go about developing a perception of risk for
something like cancer, when the relevant probabilities may range anywhere
-4 -9
between, say, 10 and 10 incidences per lifetime in the affected popula-
tion? Viewed in the context of a sequential decision making process, the
following issues arise:
1. What information should be collected?
2. How should the collected information be analyzed, inter-
preted, and translated into perceptions and thresholds
of risk? It is inevitable that any information process
designed to turn up indications of very low-probability
risks will also produce large amounts of useless, irrele-
vant, or even misleading information. In communication
theory this is referred to as "noise" within the system.
How can the weak signals pertaining to risk be filtered
out from the accompanying noise of useless information?
3. Do some organizational arrangements militate against
analyzing and focusing on potential problem areas?
That is, since persons responsible for detecting low-
probability risk must be highly sensitive and closely
attuned to the problem, how many other duties and re-
sponsibilities can they be assigned without drowning
the weak signals emanating from low-probability events?
4. Can certain administrative functions associated with
the detection of low-probability risks be replaced
at lower cost with purchased inputs? For example,
could third-party intervenors be compensated for in-
formation leading to "apprehension and conviction,"
such as triggering of formal proceedings that result
in cancelling the registration of offending pesticides?
Could this be done without, in turn, generating a flood
of useless information, along with frivolous and un-
desirable attempts at intervention?
5. What kind of feedback mechanism should exist within
the information gathering process? That is, if certain
monitoring studies indicate the existence of a poten-
tial risk, should this information be used to trigger
a more intensive information gathering process focused
on the pesticide in question, or should a basic one-
level monitoring process continue at a standard rate
of intensity as the sole provider of information?
82
-------
6. As evidence of potential risk accumulates, should the
Administrator of EPA have other mechanisms and resources,
other than triggering a formal cancellation proceeding
or convening a formal hearing, that would enable him
to accelerate the gathering and evaluation of informa-
tion pertaining to those pesticides about which doubts
have arisen? For example, should the Administrator
have resources for special ad hoc studies designed
to supplement information obtained from the regular
monitoring program?
Answers to these questions are now emerging as a result of the recently
initiated Hazard Evaluation System (HES) Program within the Office of Pesti-
cide Programs (OPP). The HES Program has undertaken a selective survey of
how hazards are evaluated in other government agencies, and has commissioned
several consulting studies. One report, A Study tc? Develop Alternative Con-
ceptual Approaches for a Pesticide Hazard Evaluation System, by Arthur D.
Little, has become available in time for the present analysis. _8/
The interrelationship between different functions connected with hazard
evaluation is shown' in Figure 1, which is a slightly revised version of a
figure appearing in the above consulting report. The most centrally connec-
ted box in this diagram is the data analysis to evaluate hazards. This func-
tion is connected to four separate functional loops: (a) one relates to data
acquisition, processing, editing, etc.; (b) another is related to the defini-
tion of hazards and changes in standards for triggering mechanisms; (c) the
third relates to construction of predictive models on pesticide applications
and possible harmful impacts; and (d) the fourth is related to identification
of hazards and adjustments (alternative application patterns) to reduce the
hazards. At present, work of the HES Program concentrates on (a) alerting
and triggering mechanisms; and (b) information/data collection, verification
and analysis.
Apart from much technical detail that will be useful in day-to-day ad-
ministrative applications, the Arthur D. Little report contains the following
conceptual highlights that have a bearing on the nature of the sequential
decison-making system employed:
* In evaluating pesticide hazards, one must undertake
a risk management decision process where the goal is
to minimize the potential for future damage. This
goal is one of suboptimization since the hazard evalu-
ation function must be embedded in a risk-benefit
system. The emphasis on process represents a concern
for adaptation to rapidly changing and newly emerging
conditions.
* The present OPP decision process operates within narrow
parameters. Hazard evaluation must be undertaken in
the frame of limited alternatives: (i) registering
for a given use; and (ii) for pesticides already reg-
istered, cancelling or reclassifying registration,
83
-------
FIGURE I.
Interrelationship Between Scientific Functional
Categories of a Hazard Evaluation System
(b)
(b)
00
START
(b)
DEVELOP PREDICTIVE MODELS
ON PESTICIDE APPLICATIONS (
POTENTIAL ENVIRONMENTAL DAMAGE
CHANGE STANDARDS 4
CRITERIA/MODELS/
TRIGGERING MECHANISM
DELINEATE/
DEFINE PESTICIDE
HAZARD (INCLUDING
STANDARDS 4 CRITERIA)
HAZARD POTENTIAL
IS ACCEPTABLE
FINISH
SIGNIFICANT HAZARD
POTENTIAL
(d)
SEEX ALTERNATIVE
APPLICATION PATTERNS
TO REDUCE HAZARD
Source; A Study to Develop Alternative Conceptual Approaches for a Pesticide Hazard Evaluation System Final Report
To the Environmental Protection Administration Office of Pesticide Programs, by Arthur D. Little, Inc. Cam-
bridge, Mass., April, 1976. Figure D.2, p. <*7; as revised by OPP.
-------
or refraining from a change. There is no option of
evaluating a pesticide hazard in relation to other
technically or economically feasible alternatives:
for example, biological pest control methods or sub-
stitution of crops that require less hazardous pesti-
cide applications. Yet "hazard potential" and "degree
of hazard" can be adequately determined only in rela-
tionship to alternative courses of action and the
values or objectives in question. _9/
* Pesticide hazards can be adequately evaluated only
in the broader framework of evaluating pest management
in its entirety. Such an evaluation would require
that "EPA explicity acknowledge the different and
sometimes conflicting values that must be balanced
(economic, food, or ecological considerations) in
meeting the best interests of the American public,"
and that it evaluate hazards "not in isolation but
in comparison with, or within the context of biolog-
ical, ecological, integrated, and other pest control
slternatives." 10/
The evaluations of carcinogen hazards. Concurrent with the evaluation
of pesticide hazards in general, EPA has initiated significant activity in
evaluating carcinogen hazards that arise in all areas of concern to the
agency. An internal draft document, "Interim Procedures and Guidelines for
Health Risk and Economic Impact Assessments of Suspected Carcinogens" 11/ is
a joint effort by several working groups, prepared in response to a special
request by the Administrator. Highlights of this document are:
* It points out that evidence has been accumulating to
the effect that the no-threshold concept widely accept-
ed for cancer induction by ionizing radiation is also
applicable to chemical carcinogens.
* It defines two key decision points concerning carcino-
gens: first, a decision whether a particular substance
constitutes a cancer risk; second, what, if any, regu-
latory action should be taken to reduce that risk.
* It posits that the evidence for carcinogenicity will
be viewed "in terms of a warning signal, the strength
of which is a function of many factors including the
quality and scope of data, the character of the toxico-
logical response, and the possible impact on public
health." 12/
* It recommends establishment of a Cancer Assessment
Group of senior scientists within the Environmental
Protection Agency, with liaison to the Department of
Health, Education and Welfare. All data on potential
85
-------
carcinogenicity, including data on suspect pesticides,
are to be reviewed by this group.
Sequencing of cancellation and suspension. Turning our attention from
the hazard monitoring and evaluation stage to the cancellation and suspension
proceedings, it is recalled from Chapter 6 that different criteria are used
for the various decisions involved. A Notice of Intent to Cancel triggers
the most expensive and time-consuming information gathering process available
to the Administrator. The ultimate decision that can result from a cancel-
lation process is, in keeping with its high cost in terms of both money and
time, the most drastic: permanent cancellation of the pesticide's registra-
tion. At the same time, issuance of this notice is triggered by the lowest
threshold of perceived risk. The only requirement is for some scientific
evidence that a pesticide (or its metabolites) may be carcinogenic.
If the Administrator also perceives that the risks associated with a
pesticide are such that continued usage presents an immediate hazard, then
a Notice of Intent to Suspend will initiate an expedited and less expensive
information-gathering process. The most restrictive action that can result
from this information-gathering process is a temporary suspension of registra-
tion pending the outcome of the cancellation proceeding.
Thus the cost of the information process used to reach a decision about
suspension or cancellation varies in proportion to the gravity of the actions
that can result from the process. The cost of the information-gathering pro-
cess triggered by a suspension or cancellation notice varies inversely with
the urgency of the perceived risk; i.e. , those risks perceived to be more
urgent trigger a less costly information gathering process. Viewing the
existing process as one of sequential decision making raises the following
questions.
1. If one were designing an optimal decision process,
would it be appropriate for the less urgent risks to
lead directly to the most intensive, time-consuming
and expensive information-gathering process? As an
alternative should an expedited and less costly sus-
pension hearing always precede a cancellation hearing?
On the basis of such a hearing the Administrator could
make a concurrent decision whether to suspend and/or
proceed with a full-scale cancellation hearing. The
public record assembled during a suspension hearing
could then be used to facilitate any subsequent can-
cellation hearing. If the suspension hearing preceded
a cancellation action, it could be modified by new
legislation to lead to a less drastic terminal action
than eventual cancellation. For example, some uses
of a pesticide could be restricted with regard to the
amount injected into the environment; other uses could
be made conditional on special crisis situations, such
as a predicted cutworm infestation of the corn crop
in a particular area. This opens up a promising di-
rection for new policy initiatives.
86
-------
2. Instead of going directly into a formal suspension
hearing, should the Administrator perhaps have avail-
able to him some other intermediate decision process
(it could be adversary or non-adversary) in which manu-
facturers and other interested parties might be invited
to participate? At present, certain alternatives do
exist, but each one has certain limitations:
a. Issuance of a Rebuttable Presumption Against
Registration (RPAR) initiates an information
gathering process, but it applies only to pesti-
cides awaiting registration or reregistration.
Moreover, issuance of an RPAR does not involve
a lower level of hazard than denial of regis-
tration; it serves simply as a way station to
the latter, more drastic step. 13/
b. The 1975 legislative amendments to FEPCA 14/
provide for consultation with the Secretary of
Agriculture and with a specially created Scien-
tific Advisory Panel prior to issuance of a no-
tice of intent to cancel or reclassify the reg-
istration of a pesticide. This process tends
to serve only as a way station to stronger ac-
tion rather than as an intermediate decision
process that operates at lower levels of per-
ceived risk.
c. Convening of a formal hearing under Section 6(b)
(1) of the law, to help the Administrator deter-
mine whether to proceed with a formal cancel-
lation notice.
The two processes both lack an appropriate, milder form of
remedy — such as restriction of the total amount of the
pesticide that can be released into the environment in any
given area, over a given time period — short of suspension
or cancellation. Instead, these "intermediate" processes
either abort altogether or proceed inexorably toward a can-
cellation notice, a remedy perhaps too drastic in relation
to the perceived level of risk.
The end result of any regulatory decision process concerned with carcino-
genic risks must allow for cancellation of a pesticide's registration as one
course of action. This is the most complete and permanent remedy available.
It is appropriate, therefore, that the most intensive information-gathering
process should precede invocation of this remedy. Also, due process of law
under our constitution necessitates that this final stage be conducted as an
adversary process, with the advocates of continued registration given a fair
hearing. Abstracting from this legal perspective and viewing the existing
process as one of information gathering and sequential decision-making raises
some interesting theoretical issues. The sequential decision process dis-
87
-------
cussed in this section'embodies, under the law, a component of risk-benefit
analysis, the subject to which we now turn.
Risk-Benejit Analysis
Congress has made the balancing of risk and benefits the key criterion
for decisions along the various lines of the existing sequential decision
chain. This section deals with some of the theoretical issues raised by
this fundamental risk-benefit issue.
Estimating magnitude of the risk. When the probability of an occurrence
-4 -9
is small — somewhere in the range of 10 to 10 , for example — it is ex-
tremely difficult to develop an estimate which one feels is correct within
one or two orders of magnitude. It is entirely possible that there may not
even be one observed occurrence of the event over a period of many years. In
the nuclear fuel cycle, for example, no nuclear generator has ever experi-
enced a core meltdown, yat the Wash-1400 Study had developed a decision-tree
approach for estimating the probability of a core meltdown. It is not neces-
sary to agree with the estimated probabilities in order to appreciate that
the estimation procedure is based on a method which, although extremely com-
plicated and somewhat debatable, is at least sufficiently objective to permit
analysis and criticism by others who have an interest in the issue.
In the area of pesticides and the risk of cancer, even the beginnings
-4 -9
of a comparable estimating procedure are lacking. While 10 and 10 are
both small numbers, there is a vast difference between the two. One expert
witness during the chlordane/heptachlor suspension hearing noted that if the
probability of developing cancer from pesticide residue were 10 incidences
per lifetime, the application of this probability to a population of
200,000,000 would result in 20,000 additional deaths from cancer (rather than
from other causes). This calculation, while clearly correct, was simply put
forth as an arithmetic exercise by the witness, who was not willing to state
that in his professional judgment the best estimate of the probability was
-4
indeed 10
The preceding exercise can be extended to state that a probability of
10 implies only 200 additional cancer deaths in a population of
o
200,000,000, and a probability of 10 implies only 2 cancer deaths. In view
of the fact that over 50,000 die each year in automobile accidents, 2 deaths
would not rank high on a list of social hazards, whereas 20,000 deaths would
rank quite high. The issues which this arithmetic exercise raises are:
What is a good professional judgment concerning the correct probability? On
what is such judgment based? Can one develop a more objective means for es-
timating the probability of developing cancer caused by pesticide residues?
A risk-benefit model for pesticide-induced cancer. There appears to be
a rather sharp contrast between the expanding body of detailed data relating
to pesticide residues, and the lack of any formal risk-benefit model for
88
-------
utilizing such data as they accumulate. The contrast with studies of atomic
risk is striking. Those concerned with risk-benefit analysis for atomic re-
actors have constructed an enormously detailed and elaborate risk-benefit
model (in the Wash-1400 study), but in the area of potentially carcinogenic
pesticides in humans, there appears to be no formal risk-benefit model what-
soever. Considerable data were introduced into the record concerning the
amount of pesticide residue which the monitoring program had found deposited
in various parts of the human body, such as the liver, spleen, or adipose
tissue (see Chapter 7 for more details). Yet there was, and is, no complete
model for assessing the risk of cancer, a fact which, at least indicates the
extent of the gaps in our knowledge. To be sure, the various scientists who
testified for both the government and the manufacturer of chlordane and hepta-
chlor were all very careful not to overstep scientific boundaries of knowledge
and draw any unwarranted conclusions from these data. The point is, even an
expedited hearing accumulates large amounts of highly technical data, and
in a full cancellation proceeding still more data are submitted. At present
there appears to be no method (other than common sense) whereby the judge can
systematically assess such volumes of data. A formal model, to take account
of and put into perspective the many branching possibilities would appear to
be a useful contribution (such a model would be similar, in broad outline, to
the Wash-1400 model).
We suggest that a multidisciplinary effort to specify and develop a pre-
liminary but formal and detailed model would represent a worthwhile contribu-
tion, even if the first such model were highly tentative. As with the Wash-
1400 model, there exists a greater number of potential branches to be speci-
fied and quantified. If nothing else, such a model could help focus atten-
tion on a number of interesting and important questions which biological
scientists might otherwise overlook. The following are of particular inter-
est: (1) How long do various residues and their metabolites persist in
humans, i.e., how long is the half-life of the various residues in human
tissue? (2) How powerful a carcinogen is each of the various residues and
its metabolites? Is it feasible and worthwhile to construct a scale which
measures the relative strength of various carcinogens? (3) Do residues have
a tendency to concentrate in certain organs or do they spread uniformly
throughout the body? (4) Does the concentration in various organs make any
difference? That is, do some organs develop cancer more readily than others?
Only the answers to questions such as these, obtained under varying circum-
stances of exposure, will clarify the carcinogenic hazards of given dosages
to persons with different life circumstances, by age, sex, degree of physical
activity, general health and exposure to other harmful substances and condi-
tions.
Data from animal experiments. At present, owing to the infeasibility of
epidemiological studies 15/ (discussed in Chapter 7) and the absence of any
formal risk-benefit model of cancer in humans (discussed above), animal ex-
periments are the only way of assessing the carcinogenic hazards of toxic
substances such as pesticides. The theoretical problems raised by such ex-
periments have been discussed exhaustively elsewhere. The chlordane-
heptachlor suspension hearings cited some of this evidence, 16/ which was
presented on the basis of both published and unpublished materials and further
developed by witnesses on direct and cross examination. The main issues are:
89
-------
* Statistical extrapolation from high dosages needed
to get measurable results in animal experiments, to
the much smaller doses that pose environmental hazards
in reality.
* The feasibility of measuring such effects directly
by greatly increasing the number of experimental ani-
mals used.
* Interaction between separate carcinogenic substances
that may increase the hazard far more than in a simple
additive way.
* Extrapolation of results from experimental animals to
man.
* The social distribution of costs and benefits.
The main unresolved problems are summed up in the following passage,
representing the views of researchers who have done some of the historical
work in the field: 177
Where does all this leave us?
It leaves us able to develop rather good animal data at
dose levels that do not really interest us. That is a
first-class highway that takes us where we do not want
to go. It leaves us unlikely to be able to develop good
data at "realistic" doses. To extrapolate animal results
to man exposed at these "realistic" doses today requires
assuming a mathematical model of dose-response in the
animal and conservative use of this model. Then we have
to jump from one species to another in ignorance of the
terrain of the landing site, i.e., the many species dif-
ferences. However, we will have knowledge of some impor-
tant special similarities and are just beginning to under-
stand some of the implications of the arithmetic. We have
begun to see that there are few, if any, good ways of
totaling the costs or computing the benefits. Cost-benefit
may be another blind alley.
Tomorrow and the next day we must do the appropriate re-
search on species differences in metabolism and in the
mathematics of the modeling and extrapolations — as a
minimum. The socially related issues, such as what is
an acceptable risk, what are the costs, what are the bene-
fits, must be discussed in the open, freely. This implies
recognizing that someone's costs may be someone else's
benefits. (Our medical costs are our physician's source
of living.) The inputs to the cost-benefit algebra are
not well worked out. Our ways of working must include
the adversary approach as well as the pleasanter way of
90
-------
cooperation. And today, we must get to precautionary
decisions for man's safety — inadequate as they are.
The gap between dosage response data and a complete system model. The
previous section detailed the uncertainties of assessing carcinogenic risk
based on data describing the biological response to known dosages under con-
trolled circumstances. There is, however, a broader problem, of which dosage
response is only a part. Given the discharge of specified amounts of toxic
substances into the environment, we need to know the answers to a series of
questions such as: (1) How long do pesticide residues and their metabolites
persist in the environment under various conditions? e.g., in the soil? in
runoff into streams, lakes, estuaries and the ocean? in the air? in plant
and animal tissues along the various links of the food chain? (2) How are
pesticide residues and their metabolites distributed by various convection
routes from their points of entry into the environment? (3) Where and under
what conditions are different concentrations of potentially carcinogenic sub-
stances generated? (4) What are the routes of entry of such substances into
normal tissues and the resulting concentration in the tissue?
This is the point at which dosage response information becomes extremely
relevant. We suggest that future work along these lines might attempt to
throw more light on specific response mechanisms, with the ultimate aim of
developing a complete system model which would (1) make good use of all
available data, and (2) demonstrate the extent of the gaps in our existing
knowledge.
Alternative pesticides vs. alternative uses of land. Cancellation or
suspension proceedings are typically initiated against one or two pesticides
at a time. This means that in practice the risk-benefit analysis is applied
against the assumption that the use of all other pesticides is unaffected.
There is thus an underlying confidence that subjecting potentially dangerous
pesticides to analysis, one or two at a time, and taking action where war-
ranted, will make the situation better, rather than worse. Because of the
way in which the analysis is applied, this assumption may be incorrect, for
the following reasons.
The law stipulates that the criterion for cancellation and suspension
is an "unreasonable adverse effect on the environment," which is defined to
mean "any unreasonable risk to man or the environment, taking into account
the economic, social, and environmental costs and benefits of the use of any
pesticide." _18_/ Beyond this statement, no particular risk-benefit model is
specified, and in the case of chlordane/heptachlor the Administrative Law
Judge did not stipulate that he was using any particular model or that he
was looking for any particular evidence. Both sides presented evidence on
the risk, costs, and benefits as they saw fit. Despite the absence of any
particular model specification, and from the admissibility of isolated indi-
vidual pieces of evidence in the administrative law proceedings, it is never-
theless possible to infer an implicit model that was being employed. Two
factors stand out.
First, all the major uses and applications of chlordane and heptachlor
were considered in some detail; alternative pesticides, effective against
91
-------
the same pests, were explicitly brought into the analysis. The central
thrust of this analysis, however, was on the cost and effectiveness of other
pesticides: the cost per application and the number of applications neces-
sary in order to achieve the same effect as chlordane or heptachlor. The
possible problem of residues and potential carcinogenicity of other pesti-
cides was not raised as an dissue, unless the other pesticides were already
the subject of a cancellation or suspension proceeding (e.g., mirex, which
is a substitute pesticide for some uses, was the subject of a concurrent can-
cellation proceeding). Thus the costs and benefits, but not the risks, of
other pesticides were being compared with the costs, benefits and risks of
chlordane and heptachlor.
The chlordane and heptachlor case shows that EPA's decision process
might well be described as "sequential suboptimization." There is an impli-
cit faith that this step-by-step process will in time lead to an optimal sit-
uation. This assumption is not altogether unreasonable, especially if the
decision process is reversible; that is, if a previously banned pesticide can
be reconsidered on its merits at a later date. In theory such a reversibil-
ity could occur. Given the ponderous nature of the present process, however,
one wonders whether subsequent reconsideration is likely ever to occur.
Second, partly as a result of the extensive discussion about the avail-
ability and effectiveness of substitute pesticides, there appeared to be a
tendency to evaluate the suspension of chlordane and heptachlor only under
the limiting assumption that farmers would continue to grow the same crops,
and would use alternative pesticides or pest control procedures. Some evi-
dence was entered into the record concerning other crops that could be grown
with no pesticides whatsoever, or with entirely different pesticides designed
to control other pests (e.g., the corn cutworm, against which chlordane and
heptachlor are effective, would not be a problem if the farmer switched to
something like soybeans or sugar beets). The Administrative Law Judge gave
some consideration to these alternatives in his recommended decision. The
lengthy discussion accompanying the final order by the Administrator, however,
while discussing at some length the availability of other pesticides for the
same crops, completely disregarded the possibility of growing entirely dif-
ferent crops on land currently treated with chlordane or heptachlor. 19/
The underlying model specification was thus, predominantly, one of sub-
system — or, in economists' terms, partial equilibrium — analysis. In con-
sidering the effects of an intervention in the operation of market forces,
attention centered on direct effects, to the almost complete neglect of indi-
rect effects brough about by second and higher-order adjustment processes.
There is no need to review here the voluminous literature that exists on
risk-cost-benefit analysis. 20/ The theoretical and practical issue is
whether EPA should develop a more formal and sophisticated risk-benefit-cost
model for use by its Administrative Law Judges and attorneys for both sides.
Such a model would specify the treatment of other alternatives like those
discussed here.
The issue of "horizontal" equity. As noted previously in Chapter 3 (see
page 16), there is a widespread argument for neglecting distributional effects
in benefit-cost analysis on the basis that the effects of the particular
92
-------
project under consideration are distributed in the population independently
of the marginal social value of a dollar. This argument, which is equivalent
to assuming that there are no distributional effects, refers to what is some-
times called "vertical" equity.
The issue raised by pesticide cases like chlordane/heptachlor relates to
the income of a particular group of farmers who depend heavily on a specific
pesticide for the success of their crops. 2jV The literature on cost-benefit
analysis contains numerous discussions on compensation principles and the
willingness of individuals in society to pay others for actions which benefit
them. The fact is, however, that the law currently provides no compensation
whatsoever for those who may be adversely affected by a regulatory decision.
Issues raised by this state of affairs are:
1. Should there be some form of compensation under cer-
tain conditions? If so, should such compensation
be based on an inherent general or limited "right
to pollute?"
2. If compensation is warranted, how much should be
paid, and what principle (s) of compensation should
be used to determine the amount?
3. If one holds that those who are adversely affected
should not be compensated, what is the theoretical
justification for this position?
In the chlordane/heptachlor suspension case the final order of the Ad-
ministrator made some allowance for the hardships which suspension would
cause by (1) permitting stocks in existence to be used and (2) postponing the
effective date of suspension for use on corn.
Monitoring and enforcement costs. As noted above, administrative law
proceedings do not incorporate any kind of formal cost-risk-benefit model.
An implicit model can be inferred to a degree, from the testimony offered.
It is interesting that in the case of chlordane/heptachlor the final order
issued by the Administrator called for restricted use of both pesticides,
rather than a complete ban. Presumably, therefore, products containing chlor-
dane and heptachlor will continue to be manufactured and sold. Some enforce-
ment of the selective restrictions may thus be necessary. Nowhere in the
record, however, were the cost, problems, or efficacy of enforcement dis-
cussed. A number of issues are raised:
1. How much is budgeted for enforcement and policing
of pesticide regulations and restrictions?
2. How effective is enforcement and policing? i.e., to
what extent will restrictions like those contained
in the Administrator's order be, in fact, observed?
3. Comparing a complete ban (which is relatively easy
to enforce) with an order for restricted use, what
93
-------
additional costs for policing and enforcement are
implicit in the restricted order?
4. What implicit assumptions regarding policing and
enforcement were made by the Administrator when he
decided to issue an order for restricted use? How
valid are these assumptions?
Dynamic Considerations
The entire regulatory legislation, described earlier in terms of the six
stages of the existing sequential decision process, is focused on chemical
pesticides already in existence and on new chemical pesticides which manu-
facturers elect to submit for registration. This six-stage process is not —
and should not become — a thing unto itself, as are so many of our other
regulatory processes. The pesticide regulatory process is part of and is
inseparable from our total technology.
The real world is not static but dynamic and the model used for ana-
lyzing the total regulatory process should also be dynamic. We should, at a
minimum, enumerate specific dynamic considerations and then analyze whether
the approach to decisions under risk adversely affects the dynamic considera-
tions. Some relevant dynamic considerations are the following:
1. Development of new chemical pesticides. Does the regu-
latory process impede or inhibit the development of new
and better pesticides? Also, will it in part redirect
R & D toward desirable defensive research and the devel-
opment of less hazardous substitutes?
2. Entry of new firms. Does the regulatory process impede
the entry of new firms into the chemical pesticide
business, thereby limiting or restricting competition?
3. Development of other pesticide control techniques.
Does the regulatory process have the effect of direc-
ting research and development towards other, non-
regulated pesticide control techniques?
4. Adoption and implementation of integrated pest control
methods, and other pest practices. What is the effect
of the regulatory process on the development of ecolog-
ically sound land utilization, food production and
energy conservation methods at the level of society as
a whole? Does it move us in the direction of social
optimality in the large, or does it bog us down in
penny-wise and pound-foolish arguments over severely
limited options?
The discussion of such dynamic considerations raises a series of ques-
tions that are more general than the theoretical issues we have been focusing
94
-------
on thus far. These questions are inseparable from the broader social policy
alternatives with which this study is concerned.
Summary
The basic legislation that established the present regulatory framework
has been described, but there has been no attempt within the scope of this
study to investigate or analyze the political bargaining process that accom-
panied the enactment of this legislation. The registration portion of the
regulatory process has also been described briefly but not analyzed. The
focus of this case study is on the adversary proceedings which take place
when the registration of a pesticide is challenged. Using the chlordane/
heptachlor suspension case as our benchmark, this chapter has reviewed alter-
native ways of making social decisions about continued uses of a pesticide
under conditions of risk and uncertainty.
In particular, market based approaches to decision making have been
reviewed and contrasted with the present regulatory approach. The major
challenge facing any decision process is the substantial lack of information
regarding the existence of pesticide residues in both the environment and the
human biomass. Basic information and knowledge in this area can be generated
only by continuous monitoring and research efforts: it will not result from
the usual interplay of market forces. Hence government-sponsored monitoring
and research on pesticide residues is seen to be essential.
Once the basic monitoring information is in hand, alternative approaches
to social choices are available. The two major decision approaches discussed
here — market and regulatory — both suffer from a lack of clear signals as
to what action ought to be taken on the basis of the fragmented information
available. The chlordane/heptachlor case raises a substantial number of un-
resolved theoretical issues, all of which provide fertile ground for addi-
tional research aimed at improving the existing decision process.
In the next chapter, we will examine the pesticide decision process in
a broader context. In Chapter 10 this process will be evaluated in terms of
the criteria developed in Part I of this study.
Notes to Chapter 8
1. For an elementary and highly readable exposition, see Robert L. Heil-
broner, The_Economic Problem (Englewood Cliffs: Prentice-Hall 1972).
The standard exposition at an advanced level is Gerard Debreu, Theory of
Value: On Axiomatic Analysis of Economic Equilibrium (New York: Wiley,
1959). For a critique, see Janos Kornai, Anti-Equilibrium; On Economic
Systems Theory and the Tasks cf Research (New York: American Elsevier
1972).
95
-------
2. This means that, contrary to the basic assumption of the theory that
organized markets exist at all times for all commodities, markets "fail
to exist" for part of the consequences of some actions. In other words,
the theory assumes that all economically significant influences within
the system are transmitted between participant firms and households in
the form of competitive commodity exchanges; yet in fact, some of these
influences "fail" to appear in the form of commodities traded in competi-
tive markets. External economies and external diseconomies are, then,
economic effects which bypass market channels.
3. See: Kenneth A. Arrow and R. C. Lind, "Uncertainty and the Evaluation
of Public Investment Decisions" American Economic Review (June, 1970);
Jack Hirschleifer, "Investment Decisions Under Uncertainty: Application
of the State Preference Approach" Quarterly Journal of Economics (May
1966).
4. See K. Lancaster and R. Lipsey, "The General Theory of the Second Best"
Review of Economic Studies, Vol. XXIV (1) No. 63, pp. 1-18.
5. Recent studies of products liability insurance have shown, however, that
less constructive adaptations may also occur. For example, firms may
break up their operations which are at risk into small legally indepen-
dent units, each of which can be abandoned to bankruptcy in the face of
a large lawsuit. This practice is already common in the case of New
York City taxicab fleets, where each cab, for legal purposes, franchises
as a separate corporation.
6. The risk may be uninsurable at any reasonable level of premiums, as has
recently happened to medical malpractice insurance. The so-called "prod-
ucts liability problem" arises from the fact that the doctrines of
strict liability and implied warranty create tort liability exposure to
individual firms that in extreme cases can exceed their net worth many
times over. For example, manufacturers of machine tools whose products
are still in use thirty or forty years after the original sale — and
this is quite common — find that the new legal interpretations saddle
them with a huge backlog of tort liability exposure that cannot possibly
be handled under commercial insurance practices. Even in less extreme
cases, the rates have risen sharply and insurance sometimes becomes un-
obtainable.
7. Majone, Giandomenico, "Choice Among Policy Instruments for Pollution
Control" Policy Analysis (1976) pp. 589-614.
8. U.S. Environmental Protection Agency, Office of Pesticide Programs, A
Study to Develop Alternative Conceptual Approaches. . ..
9. U.S. Environmental Protection Agency, Office of Pesticide Programs, A
Study to Develop Alternative Conceptual Approaches. . ., p. 41.
10. U.S. Environmental Protection Agency, Office of Pesticide Programs, A.
Study to Develop Alternative Conceptual Approaches. . ., p. 43.
96
-------
11. U.S. Environmental Protection Agency, Interim Procedures and Guidelines
for Health Risk and Economic Impact Assessments of Suspected Carcinogens
(Washington, B.C.: May 13, 1976).
12, U.S. Environmental Protection Agency, Interim Procedures and Guidelines.
_. ._, p. 4, emphasis added.
13. See Chapter 6, subsection: "Current legislation concerning pesticides,"
Item l(c), p. 49.
14. 94th Congress, Public Law 140, 89 Stat. 751, 7 U.S.C.S. Sects. 136-136b,
136p, 136w, 136y, November 28, 1975.
15. See Chapter 7, Section: "Evaluating the Risk of Cancer," second para-
graph, p. 64.
16. See references cited in Chapter 7, Notes 13 and 17.
17. Schneiderman, "From Mouse to Man. . ."
18. See Chapter 6, subsection: "Current legislation concerning pesticides,"
2nd paragraph, p. 48.
19. U.S. Environmental Protection Agency, Suspension Proceedings Before the
Commissioner, Decision of the Administrator on the Suspension of Hepta-
chlor/Chlordane, and U.S. Environmental Protection Agency, Before the
Commissioner, Recommended Decision, p. 86.
20. Some of the more important references used in this study are included
in the bibliography.
21. Pesticide decisions inevitably impact certain geographical areas and
parties. To the areas and parties affected, these impacts are often
major. At the same time, aggregate impacts — that is, impacts on
commodity prices and consumers — may be quite small. Thus, issues of
horizontal equity are pervasive and inherent in pesticide decisions,
while issues of vertical equity may be so minor as to be almost non-
existent.
97
-------
9_. Broader Issues Concerning the
Pesticide Decision Process
The discussion in Chapter 8 emphasized those aspects of the decision
process that fit into one particular overall view of reality: it surveyed
the approaches of neoclassical theory to the problem of decision-making under
risk, analyzed some aspects of EPA's sequential decision-making process, and
reviewed the risk-benefit approach embedded in this sequential decision-making
process. For lack of a better label, we shall designate that viewpoint as the
"optimization" paradigm. I/
The total decision process is somewhat broader than the methodology for
optimizing costs, risk, and benefits. Other approaches to the pesticide de-
cision process which are at least as important, perhaps more so, are examined
in this chapter. The optimization paradigm of Chapter 8 will be compared and
contrasted with one alternative view of reality, the "legal" paradigm. This
introductory discussion puts the optimization paradigm in perspective and
highlights the importance of using a socially acceptable process to make im-
portant decisions and social choices (for more extensive discussion on this
particular point see Part I of this study).
The legal process as a vehicle for decision-making under risk has, in
recent times, been criticized by many scientists who feel that better ways to
marshal and weigh scientific evidence can and should be developed. In view
of this concern by the scientific community, this section includes a brief
examination of some alternative legal procedure which might be more acceptable
to them, as well as a brief rationale for each suggested alternative. The
purpose of this discussion is to develop a broader understanding of the de-
cision process as it relates to pesticides. The concluding chapter will then
evaluate the existing decision process in terms of the criteria presented in
Part I of this study.
Alternative Paradigms for Decision-Making Under Risk
The process of arriving at social decisions in situations involving low-
probability high-consequence outcomes involves people from a number of disci-
plines and backgrounds; each brings his own view of reality to the decision
process. In the legislative part of the process, exemplified by enactment of
the FIFRA legislation, an important view of reality is that held by the poli-
tical scientist, or what might be called the "political" paradigm. 2J This
viewpoint is downplayed here only because this case study is focused on that
part of the decision process which follows enactment of the legislation.
99
-------
Current legislation requires that the Administrator of EPA take account
of the risks and benefits of continued use when deciding whether to
continue registration of a pesticide. The law does not require that pesticide
cases specifically employ an optimizing cost-risk-benefit model, but intent
seems implicit. At the same time, the FIFRA legislation also requires that
the risks and benefits be considered within the framework of an administra-
tive law proceeding. The optimizing and legal paradigms thus represent the
two major approaches embodied in the current pesticide decision process.
The optimization paradigm. This paradigm might best be characterized as
oriented to the modeling of current decisions about future situations, while
the legal paradigm is oriented more to a process for resolving past disputes
and incorporates a strong emphasis on precedent and authority. As applied to
risky decisions, the optimization paradigm attempts to identify whatever in-
formation is available for dealing with a problem, and then within the con-
straints of that information hopes to find a way to reach the best possible
decision. Of course, this decision need not be optimal in its actual outcome
when the outcome itself is probabilistic; it is required to be optimal only in
the sense that, under the given circumstances, the expected payoff is being
maximized. 3J
Approaching problems within the context of this optimization paradigm
requires that all major factors pertinent to the final decision be quantifi-
able. The optimization approach then operates within the broad outline of the
cost-risk-benefit calculus. Important nonquantifiable factors, such as the
dynamic considerations discussed in Chapter 8, are typically not included in
the formal model; more often than not, they are deemphasized or omitted al-
together from the analysis. The cost-benefit model can theoretically be ex-
panded to include a comprehensive systems approach, as in a total ecological
system, but the model that is actually used for a specific problem often rep-
resents a subsystem of some broader unspecified system. Thus specific appli-
cations of this optimization approach may vary quite widely, from one case to
the next, depending upon the viewpoint of the analyst and the possibilities
of measurement.
Risk analysis may range from, an elaborate and extensive formal model like
that contained in the Wash-1400 study kj (the Rasmussen Report) to no more
than a professional judgment by a technical expert. The optimization approach
attempts to arrive at a single solution, based on maximizing one specified ob-
jective function. While multiple objectives may be contained in a composite
objective function, for the most part the optimization paradigm attempts to
offer a unifying approach. It is oriented neither to conflict, especially
deep-seated conflict, nor to the resolution of conflict.
The legal paradigm. The legal outlook, in contrast, embraces the reali-
zation that a given set of facts can be interpreted from at least two (or
more) contradictory angles. This approach is more open to perceiving a prob-
lem in its entirety, albeit in a less formalized setting. As applied to de-
cisions involving pesticide risks, this approach recognizes the tremendous
information gap which characterizes any individual decision, and does not
necessarily aim at making individual decisions in some sense "perfect". It
is concerned more with what can be reasonably said to be true about a partic-
100
-------
ular situation than with what the abstract truth of that situation actually
might be. In other words, it concentrates on pesticide decisions as a stream
of human activity involving much uncertainty, many interests in partial or
total conflict, many independent or conflicting value orientations among par-
ticipants, and the impossibility of making, in any sense, a "perfect" decision
in a given case. 5/ In each specific case the legal process reaches a deci-
sion by applying known rules of law to facts admitted under the established
rules of evidence (and only to those facts).
An essential consideration is the formation of a social consensus that
emerges as a result of the tugging and pulling of the long-term, repetitive
legal process. At the end of this process, individual participants may feel
unsatisfied with particular outcomes, but they will usually also feel that
they have had their "day in court" and therefore are generally more willing to
accept a somewhat adverse outcome. If the legal approach is established with-
in an executive agency, as was done with pesticide regulation, the ultimately
political nature of any regulatory decision making process is stressed,
therefore necessitating the clear identification and political accountability
of the final decision maker. _6/
Comparison of the two viewpoints. The optimizing and legal paradigms
both represent cognitive approaches to find the "best" choice in a given sit-
uation, but there are some fundamental differences between the two points of
view. Persons trained in these contrasting ways of thinking may have diffi-
culty communicating with one another. Fine lawyers sometimes feel uncomfort-
able resting their cases on risk-benefit calculus. Conversely, many scien-
tists — and this emphatically includes some economists — are not comfortable
thinking in terms of social process and conflict rather than in terms of
optimizing choices. TJ
Although the preceding description stressed differences between the op-
timizing and legal approaches, these two viewpoints are not mutually exclu-
sive, nor do they represent opposing viewpoints. Instead, their differences
make them complementary, and the present decision process can be viewed as an
attempt by Congress to combine the two paradigms into a unified decision pro-
cess. That is, the optimization paradigm (cost-risk-benefit analysis) is
embedded within the adversary process used for cancellation and suspension
proceedings. This legal approach in turn reflects our political process,
which provided the authorizing legislation.
An appraisal of EPA's sequential decision-making process must be capable
of embodying not only these two points of view, but others as well. 8J Pro-
cedures which might otherwise seem hard to justify suddenly come to life when
one switches from one point of view to an entirely different one. This is
particularly important in light of Congress' recent concern for opening up
the EPA decisionmaking process to more public participation, especially to
agricultural interests _9/ which have seen many recent EPA decisions go against
them.
Another important viewpoint affecting the legal process used in pesti-
cide cases is the scientific viewpoint. This might best be characterized as
oriented to the accumulation of knowledge and the drawing of logical infer-
101
-------
ences where possible, coupled with the abstention of inference and retention
of an open mind when inferences can not be supported by facts and experimen-
tal evidence. 10/ The scientific viewpoint, with its strong emphasis on ra-
tionality, is not really a social decision process; its presence is nonethe-
less strongly felt in pesticide cases.
An article by Philip L. Spector, "Regulation of Pesticides by the Envi-
ronmental Protection Agency" ll/ contrasts the views of scientists on the one
hand with those of lawyers on the other hand — all within EPA. Spector
avers that lawyers, clustering around the Office of General Counsel, are in
general more environmentally activist than scientists associated with the
Office of Pesticide Programs, but our own interviews within the Agency found
strong dissent with this point of view.
Specter's allegation is certainly too broad. Scientists do appear on
different sides of environmentally relevant issues, but as a group they cannot
be characterized as less committed to the environmental protection principle
than lawyers. Spector bases his point on the reluctance of scientists to
commit themselves to the defense of the hypothesis that a particular pesticide
is harmful, unless there is compelling evidence to support such a view. He
fails to state, however, the corollary that scientists are equally reluctant
to commit themselves to the hypothesis that a pesticide is harmless unless
there is evidence to support that particular statement. In between there is
an enormous range of cases where the evidence supports neither the original
statement nor its converse. Scientists — as scientists — are neutral be-
tween hypotheses; they judge only the quality of support.
Aside from Spector, a great deal has been written about both the scien-
tific and legal method, but to our knowledge the similarities and differences
between the scientific and legal viewpoints have nowhere been carefully
drawn. Similarities center around the fact that both processes attempt to
derive valid conclusions from factual evidence. Legal and scientific method-
ology each operate within a broad framework, yet each is capable of and at-
tempts to tackle a specific problem on its own individual merits. In this
broad sense the two processes appear complementary. But certain differences
also deserve mention. One is that development of the legal process was cen-
tered for hundreds of years around cases that pertained solely to past events,
such as crimes or civil cases like breach of contract. Thus both criminal and
civil rules of evidence were designed primarily to help determine the perti-
nent "facts" related to some prior incident or act. In other words, legal
rules of evidence were not developed for the purpose of marshalling and inter-
preting information to assess the possibility of future occurrences. 12/ In
the case of risk, however, such as pertains to pesticide decisions (and in
other areas beyond the scope of this case study), this is precisely the use to
which the legal system is being put. 13/
A second difference concerns the strong underlying orientation of the
legal process toward reaching a decision. For this reason, in the legal pro-
cess (and, for that matter, in other decision processes as well) inferences
must often be drawn from fragmentary and inconclusive data. Lawyers are
trained to prepare their case, present the best evidence available and then
drive towards a resolution of those matters that require decision, one way or
102
-------
another. The length of time required to try a major legal case does not alter
the existence of this basic drive towards a final decision. Scientific train-
ing, by contrast, leads one to reject any unwarranted inference and instead
seek more information. A truly scientific outlook motivates one to indicate
the limits of knowledge and lay out research aimed at adding to our base of
knowledge. Courts (and other decision making bodies) focus on reaching
"fair", "just", or "acceptable" decisions, one way or another, and the deci-
sion process itself typically exhibits relatively little interest in develop-
ing proposals for further inquiry. 14/
Alternative Legal Procedures^
The preceding discussion has again focused on differences while glossing
over the main similarities between the scientific and legal viewpoints. How
critical the differences are between these two viewpoints is unknown, but from
time to time a number of scientists have expressed dissatisfaction with the
way in which our existing decision processes marshal and interpret scientific
evidence on technical issues. This dissatisfaction has led some concerned
scientists to recommend the establishment of an entirely new institution,
usually referred to — for want of a better name — as the "Science
Court." 15/
It is not our purpose here to evaluate or enter into the debate over the
need for, or merits of, a Science Court. The fact that a number of scientists
have expressed so much dissatisfaction with the existing decision process
may, however, indicate a gradual but steady erosion of the ability of the
process to forge social consensus on crucial issues. It is of supreme impor-
tance that the process not lose the support of a major influential group such
as the scientific community. For this reason, the issues posed by proponents
of a Science Court need thorough airing. In particular, alleged shortcomings
of the existing legal process need to be clearly enunciated, and alternatives
aimed at overcoming any generally agreed-upon shortcomings need to be devel-
oped and evaluated. The Science Court is but one of many alternative reforms
that could be adopted. Within the scope of this case study it is not feasible
to undertake a complete critique of existing legal procedures as they are ap-
plied to decisions involving determination of scientific "facts", as in a
pesticide case. On a more limited scale, though, we have explored a few pos-
sible changes which could be made in the legal procedures used to gather and
interpret scientific data that are deemed relevant to risky pesticide
decisions.
It is recognized that EPA has no authority to change the administrative
law practices which are prescribed and followed in cases involving the suspen-
sion or cancellation of pesticides. Only Congress can change the law. The
purpose of this discussion is not to criticize EPA, but to suggest some ex-
plicit alternative procedures. These are raised as topics for discussion
only. Their mention is not intended to recommend them, nor even to indicate
that we regard them as desirable.
Revise the role of the expert witness. In a legal proceeding a person
who qualifies as an expert witness is permitted to play a somewhat privileged
103
-------
role. In particular, professional opinions of expert witnesses are allowed
to enter the trial record as evidence, whereas other witnesses are not per-
mitted to express opinions as part of the record. Thus in a trial proceeding
expert witnesses are in a position to play a differential — and perhaps cru-
cial — role. Due to the importance of this role and the potential influence
which an expert witness may have on the outcome of the case, lawyers sometimes
expend considerable effort in establishing or attacking the credentials and
credibility of proposed expert witnesses.
An interesting feature about expert witnesses is that the law does not
provide for any gradations of expertise: a person either qualifies as an
expert witness, or he does not qualify. (In pesticide cases it is typical
for a majority of the expert witnesses to be natural scientists, most typi-
cally from the biological sciences.)
This legal tradition for building up evidence about a case by use of
opposing expert witnesses is somewhat at variance with scientific tradition.
A purely scientific debate typically leaves little room for opinions (other
than theories, where theorizing is necessary). Emphasis is placed instead on
experimental procedures (controls, accuracy of measurement, etc.), findings,
interpretations of data, and replicability of the results. Further, in scien-
tific debate, anyone is free to criticize the work of others. Provided a
scientist's comments are cogent, arguments over credentials, (such as where
the person went to graduate school), are of no great importance to the scien-
tific debate. To be sure, scientists are not immune to someone else's prior
work and scientific reputation and gradations among scientists do exist,
(such as graduate student, junior and senior research staff), but a person's
status does not preclude him from undertaking significant research or from
criticizing the work of others. Many scientists do their most brilliant work
early in their careers, yet the legal system often seems to militate against
admitting or using younger scientists as expert witnesses.
Turning to our interest in the decision process, we pose the following
two questions: (1) What changes could be made in the legal process to bring
it more into accord with scientific traditions? and (2) Are such changes de-
sirable? Several modifications to the present system appear readily avail-
able.
First, pre-trial discovery procedures could require that (a) at an early
stage expert witnesses be allowed or even required to submit prepared testi-
mony containing all major evidence they wish to cite and all important opin-
ions they wish to express; 16/ (b) in general such prepared testimony should
be in a publishable format, which means that it be reasonably self-contained,
including complete citations to all references and that all charts, figures
and tables be properly labelled and have proper titles, etc.; and (c) this
prepared testimony in fact be published, with EPA acting as "the publisher of
last resort." This procedure, in keeping with the spirit of pre-trial dis-
covery procedures, would greatly reduce the role of initial direct examina-
tion of expert witnesses by the counsel who introduces the witness. The
examination would be replaced by a carefully prepared written statement,
which would be more in keeping with scientific tradition. Under this pro-
cedure, important professional opinions would not generally be introduced
104
-------
during subsequent oral presentation except, perhaps, if called for under
cross-examination. This procedure will permit all expert opinions and pro-
fessional judgments introduced as evidence to be circulated widely by both
sides, with the purpose of subjecting a person's scientific views to inten-
sive scrutiny by a wide range of interested scientists. The foreknowledge
that the opposing side is likely to circulate and debate opinions more widely,
in the scientific literature as well as before an administrative law judge,
can only serve to enhance the quality of both the evidence and the debate.
A second possible change would be to permit any witness to challenge any
opinion (or any other statement) expressed by an expert witness. The only
requirement would be that any such challenger must also be subject to cross-
examination and possible rebuttal. In other words, to challenge a statement
by an expert witness, one would not have to qualify first as an expert wit-
ness. The court would, of course, be expected to consider any reasonable
challenge or criticism on its merits. A procedural change of this nature,
while possibly at variance with legal tradition, would appear to be in keeping
with accepted scientific method. Such a change might, of course, be quite
time-consuming if firms brought in many witnesses.
If the preceding change were adopted and led to frivolous and seemingly
capricious challenges (this seems an unlikely prospect, but it cannot be dis-
counted altogether), a third possibility would be to establish a new inter-
mediate category of qualified witness, whose prerogatives would be to chal-
lenge opinions and raise pertinent questions, without being allowed to ex-
press unlimited opinions before the court.
A fourth alternative, for cases that focus on the accumulation and in-
terpretation of scientific evidence related to future outcomes — such as
pesticide cases — might be to relax substantially or even eliminate alto-
gether the concept of expert witness. The idea is that the definition of an
expert witness would have absolutely minimal qualifying hurdles, beyond which
gradations in credentials and resumes would, per se, be of no particular im-
portance. Then, hopefully, the focus of the testimony, cross-examination and
debate would be on the substance and particulars of the issue. If this re-
sult were accomplished, adversary-type proceedings concerned with low-
probabability, high-consequence matters might be conducted more in accord
with "good" scientific method.
Scientific advisors to the court. As indicated previously, cases in-
volving the cancellation or suspension of a pesticide's registration, such as
the chlordane/heptachlor case, are conducted within an adversary legal frame-
work. This framework seems particularly well suited for fact-finding and
adjudication of a prior occurrence such as an alleged breach of contract.
The central concern of a pesticide case, however, is the assessment of future
risks and costs occasioned by pesticide residues against the future benefits
that would result from continued use of the pesticide. Prior "facts" are
relevant only to the extent that they help in this assessment. Thus, while
the case represents a legal trial in one sense, in another sense it can be
viewed as a process of information-gathering and scientific/technological
assessment. This latter viewpoint deserves further discussion.
105
-------
In pesticide cases, much of the information that is brought together
(i.e., the evidence) will be of a technical nature; some of it may be highly
technical. Although administrative law judges develop a certain amount of
background and expertise in the area of their specialty, their primary train-
ing and professional experience is in the law, and they may at times exper-
ience difficulty in understanding and integrating large amounts of detailed
and highly technical information. It might be a good idea, therefore, to
enable a judge to appoint special scientific advisors to assist him in sifting
and interpreting the evidence. 17/ Other public decision-makers, be they
administrators in the executive branch or congressional committees, typically
have unrestricted access to advisors and consultants.
When regulatory agencies were first established during the 1930's, the
then-prevalent theory was that an administrative law judge could draw on mem-
bers of the agency staff if he felt that he needed professional guidance or
opinions. In light of current realities, however, this is totally infeasible.
As Specter 18/ points out, EPA's Office of General Counsel treats the Office
of Pesticide Products as if it were its "client". Consequently, the Office
of Pesticide Products is cast in the role of the prime adversary party in
challenging continued registration of the pesticide. Clearly, this aspect of
the original theory underlying administrative law practices is no longer
valid. It is against the rules of both common sense and the law for a judge
to seek professional advice from one of the two opposing parties.
Scientific advisors to the court might play other useful roles as well.
For instance, the classical adversary process depends entirely on the init-
iative of the opposing sides to determine who will be the witnesses, what
will be submitted in evidence, and what questions will be asked of the var-
ious witnesses. Advocates clearly have strong incentives to put forth their
best case, and the process has undisputed strength in bringing forth a consid-
erable amount of pertinent evidence and viewpoints. When viewed as a social
decision process, however, the system is by no means foolproof. The central
issue in pesticide cases is posed with disarming simplicity. Because it is
posed as a "yes/no" or "go/no go" decision — i.e., continue or discontinue
the pesticide's registration for a particular use or uses — at first blush
the issue appears to involve only two opposing sides, whose vested interests
cover all important aspects of the issue. But broad social issues with wide-
spread ramifications (and decisions concerning risky pesticides are included
in this category) often have more than two major facets. This means that all
important interests may not be represented during the hearings on a pesticide
case. The possibility always exists that two opponents may reach an accom-
modation at the expense of a third party. At the same time, participation in
any administrative law proceeding is time-consuming and expensive. The view-
point most susceptible to being omitted is that of the general public. Con-
sequently, when this occurs, the legal-adversary process may fail to ask
questions that, from a broad social viewpoint, are highly pertinent. A ques-
tion never asked will be a question never answered. However, if everyone
attempts simultaneously to be a free-rider, none of the free-riders will be
represented. Thus a second role for scientific advisors to the court would
be to provide the judge with lists of additional evidence which should be re-
quested and all pertinent questions that ought to be asked. Permitting
106
-------
scientific advisors to play this role would of course require a concomitant
change in procedure to permit the judge to initiate requests for clarifying
evidence and information. 19/
Publication of scientific evidence. This represents an expansion of the
the idea broached earlier in the discussion on the role of expert witnesses.
Openness of communication, including publication and peer review of all theo-
ries and significant experimental results is basic to the scientific method.
The legal method is not altogether dissimilar. Legal tradition requires all
formal proceedings to be carefully and scrupulously recorded and made part of
the public record. Moreover, in preparing a case, lawyers and their witnesses
can be expected to draw on all published material that is relevant and helpful
to their position. Thus there is considerable harmony between the scientific
and legal approaches.
Discordant notes arise from other parts of the system, however. Pesti-
cide cases will inevitably involve a number of business firms: e.g., manu-
facturers, fabricators, wholesalers and distributors. With regard to openness
of communication and sharing of knowledge, the business viewpoint stands at
sharp variance with the scientific viewpoint. In a world of imperfect know-
ledge, business firms attempt to build up an internal base of research re-
sults, know-how and industrial secrets which they use to gain competitive ad-
vantage. One result of this internalization of information by business firms
is that experimental data and findings introduced as evidence into pesticide
cases will often not have been published in the scientific literature. For
unpublished evidence it naturally follows that neither the experimental pro-
cedures nor the findings and interpretation of the data will have been sub-
jected to the usual scientific peer review process (which includes possible
attempts to replicate critical experiments). From the viewpoint of our study
of the decision process, two pertinent questions are: (1) What changes, if
any, can be made to the legal process to bring it more into accord with scien-
tific traditions? and (2) Would such changes be desirable?
The volume of unpublished scientific evidence and data presented in a
trial represents only a small fraction of the total trial record. However,
such data can play a key role in the decision process. One possibility,
therefore, would be to go beyond merely establishing a public record in a
pesticide case, and instead (a) require that all expert witnesses submit un-
published data or evidence in a publishable form; and then (b) actually under-
take to have all such evidence published, in as expeditious a manner as
possible. That is, the author/witness might be allowed to submit and arrange
for publication in a journal of his choice within, say, threee to six months,
otherwise EPA would proceed with publication. Such a change in procedure
would in no way diminish the value of a witness's evidence. To the extent
that the expectation of publication causes witnesses to prepare their presen-
tations more carefully and in more detail, such a change would seem to en-
hance both the scientific and legal value of the evidence. 20/ This change
would be in keeping with the previously discussed possibility of improving
pre-trial discovery procedures.
107
-------
Conclusion
A number of different viewpoints — e-g-> legal, optimizing, scientific
— play important roles in the pesticide decision process. A full understand-
ing of the*total process requires an appreciation of all viewpoints. Because
of the somewhat limited scope of this case study an in-depth exposition of
the strengths, values, and weaknesses of each viewpoint is not possible. The
discussion here has focused instead on certain critical differences between
these viewpoints, the purpose being to sharpen the reader's awareness of the
impact of each viewpoint on the total process.
It should also be pointed out that despite the differences discussed
here, there is a single unifying theme. Namely, each of the different groups
attempts to apply rational rules of reasoning to arrive at the "best" social
solution. Each group is trained in and relies on its own methodology. While
these may diverge at certain points, there are many key similarities.
Social choices concerning low-probability, high-consequence occurrences
are often made in an atmosphere that is emotion laden and intensely personal
for large numbers of people. Major vested interests, with huge investments
in research, plant facilities and equipment may also be at stake. Social ac-
ceptance of the process by which choices are made is of utmost importance as
we have seen from Part II of this study. Critical differences in cognitive
methodologies tend to reduce acceptability of the process by influential
groups, whereas similarities enhance acceptability. Therefore, the one re-
comendation that emerges from the discussion to this point is that more com-
parative study of the way cognitive methodologies relate to the decision
process is needed.
This chapter concludes our review of the pesticide decision process as
it now functions. The next and concluding chapter evaluates the pesticide
decision process against the eight criteria developed in Part I of this
study.
Notes to Chapter 9
1. A paradigm is said to be a mind-set underlying the interpretation of
reality. It is more fundamental than a theory, it underlies particular
theoretical approaches. It extends not only to the content of theore-
tical formulations built on it, but also to their form; to be in keeping
with a paradigm, an approach must make use of a particular methodology,
and must use particular instruments in particular ways. See Thomas
Kuhn, The Structure of Scientific Revolutions, reprinted as a paperback
from the International Encyclopedia of Unified Science 2nd Edition,
enlarged, Vol. 2, (Chicago: University of Chicago Press, 1970).
2. The "political paradigm," consists of power, power blocs, conflicts in
basic value judgments, conflicts of interest, methods for resolving con-
flicts, voting blocs, compromises, etc. The political paradigm is
108
-------
especially applicable to the process which accompanies the enactment of
all pesticide legislation. For one view of the political paradigm as it
relates to the social decision process, see Ida R. Hoos, The Assessment
of Methodologies for Nuclear Waste Management" (mimeographed manuscript,
undated), or "The Credibility Issue" (mimeographed manuscript, August 11,
1976).
3. For an elementary survey of decision theory, see A. N. Halter and G. W.
Dean, Decisions Under Uncertainty (Dallas: South-Western Publishing
Company, 1971).
4. U.S. Nuclear Regulatory Commission, Reactor Safety Study, An Assessment
of Accident Risks in U.S. Commercial Nuclear Power Plants (October 1975)
(NUREG 75/014 Wash-1400).
5. In its own way, the legal approach is also concerned with doing the best
it can, in the light of the many realities mentioned here. Thus the
legal outlook might also be dubbed an "optimization" approach, but it is
optimization in the face of severe conflict and in light of broader
considerations.
6. Many regulatory agencies, unlike EPA, are designed to be independent of
the executive branch. In both theory and practice, agencies such as the
ICC are less politically accountable than the Administrator of EPA, who
serves at the pleasure of the President.
7. All too often cost-risk-benefit models fail to treat explicitly the
range of uncertainty surrounding certain variables. Techniques do exist.
Parametric analysis, showing how different values affect the final out-
come, can be used. Rarely, if ever, does a cost-benefit model contain
a variable indicating the extent of conflict involved in an issue.
8. For instance, many scientists are specialists trained to look at rela-
tively narrow issues, and to think of optimization within narrow con-
straints. Others, such as systems analysts and ecologists, have a some-
what different mind-set because they are trained to think broadly about
interdependencies and linkages of narrower subsystems.
9. See Notes 11 and 12, Chapter 6.
10. It is possible, at least conceptually, to consider all practitioners of
cost-benefit analysis to be scientists, from one discipline or another.
But all scientists are by no means "optimizers" or practitioners of
cost-benefit analysis. The "scientific" paradigm thus differs from the
"optimization" paradigm.
11. Spector, "Regulation of Pesticides. . .".
12. An important and much-discussed aspect of the legal process is its effect
of deterring future actions deemed undesirable. But deterring future be-
havior by people and assessing the probability of future occurrences,
109
-------
especially when all parties admit that the probabilities are "low", are
not the same thing.
13. Non-lawyers sometimes express opinions to the effect that the law over-
emphasizes "procedure" at the expense of "the truth". A hypothetical
example of such an instance would be the following situation. Suppose
someone had stolen internal records from a business firm which indicated
that over a long period of time a particular product would cause biologic
malfunction of some kind. Suppose further that the only way this evi-
dence can be replicated is by long, drawn-out controlled experiments.
Evidence obtained by illegal means typically cannot be introduced into a
formal trial. Some would feel, however, that rule of evidence is inap-
propriate in a legal proceeding aimed at preventing future occurrences,
not at convicting someone for past events.
14. Thus to some scientists any decision process will appear to reflect an
unscientific outlook.
15. See, for example, Arthur Kantrowitz, "Controlling Technology Democrat-
ically". American Scientist, September-October, 1975, pp. 505-509, and
reference cited therein.
16. The procedure of requiring expert witnesses to submit written statements
in advance is already used by some regulatory agencies, such as the
Federal Communications Commission.
17. Rules and procedures governing the conduct of such scientific advisors
would have to be developed. Ex parte communication is one potential
problem. Perhaps all communication between the judge and any such advi-
sor should either occur in the courtroom (before the opposing parties),
or it should be in written form and made available to both opposing
parties.
18. Spector, "Regulation of Pesticides. . .".
19. The idea of a science advisor to the court is the "inverse" of the
Science Court concept, wherein scientists would play the role of judges,
and judges of the Science Court would have legal advisors available to
them.
20. In a formal suspension or cancellation proceeding, business firms are
not required to submit in evidence proprietary information which supports
their case. However, they typically find it in their best self-interest
to do so. After business firms voluntarily enter information into a
formal proceeding, it becomes part of the unpublished public record.
Publication of such information is a relatively modest proposal. A
major issue confronting EPA, and not touched on here, is how to reconcile
the Freedom of Information Act with proprietary information submitted as
part of the registration process.
110
-------
10. Evaluation of the Pesticide Decision Process
This chapter aims to bring together the facts of the chlordane/heptachlor
case study and the theoretical considerations presented in the preceding chap-
ters to arrive at a tentative evaluation of the entire pesticide decision
process.
In a broad perspective, it should be recognized that the cancellation and
suspension proceedings for pesticides represent only one of many decision pro-
cesses for making social choices under conditions of low probability, high
consequence outcomes. When decisions must be made involving similar circum-
stances in other areas — such as the supersonic transport, the siting of
nuclear reactors, or disposal of nuclear wastes — a process of gathering,
presenting, disseminating and weighing scientific information is undertaken
by the government. From one situation to another, the process may vary con-
siderably. Congressional committees, for example, do not use a formal adver-
sary proceeding; instead, they utilize what is referred to in legal parlance
as an "inquisitorial" proceeding. _!/ That is, persons who testify before
congressional committees are not cross-examined by opposing legal counsel, but
are subject to interrogation by members of the committee. The process of
determining who will testify before congressional committees differs from
adversary proceedings and can, on occasion, be somewhat political.
The National Academy of Sciences, while not a formal decision making
body, provides important input to the decision process by conducting investi-
gations and issuing reports. Committees of the National Academy, like com-
mittees of Congress, use inquisitorial procedures for hearing witnesses and
gathering information on a subject. However, committees of the National
Academy do not typically hold public hearings, nor do they publish the testi-
mony and proceedings that take place before their committees. Also, since
committees of the National Academy are far less political than committees of
Congress, the determination of whom to invite, or whom to permit to testify
is subject to fewer political influences and considerations.
As was pointed out in the preceding chapter, the inquisitory procedures
used by both the committees of Congress and the National Academy of Sciences
are not the only institutional alternatives available. A number of interested
and informed parties have advocated that a new type of adversary proceeding
be established. The proposed institution is presently called, for want of a
better name, the "Science Court". Advocates of this concept feel that serious
shortcomings exist in our existing institutions for gathering and weighing
scientific evidence on important issues of social choice. In general, the
Science Court would complement, not replace existing decision processes.
Ill
-------
The existence of such different ways for gathering, disseminating and
weighing scientific information, coupled with the fact that new institutions
and modifications of our existing institutions can be invented, focuses atten-
tion on the need for criteria to compare and evaluate our decision processes.
Part I of the overall study has developed one such set of criteria. In this
concluding chapter of Part II, the pesticide case study, we shall evaluate the
decision process used in the chlordane/heptachlor suspension case by reference
to the eight criteria specified in Chapter 4, Part I. For ease of reference,
these criteria are listed in the accompanying Table 5.
1. Are parties who incur the present costs adequately represented?
Parties presently at risk are all members of the general public who ingest the
pesticide residues. Additional costs are born by anyone who has developed or
will develop cancer as a result of ingesting pesticide residues. Those who
develop cancer, together with their families and associates, bear all the
human suffering which accompanies the disease. Some of the monetary burden,
namely the cost of treatment covered by health insurance, is spread directly
among the general public. However, since everybody is at risk, to quite a
large degree it is the general public that is affected by the harmful resi-
dues. In the chlordane/heptachlor case the interest of the general public
was represented by EPA from within the public sector and by the Environmental
Defense Fund from the private sector. No health insurer joined the case to
represent the public interest, and no other government agency joined the case
on the side of the public. In other words, the Secretary of HEW, the Surgeon-
General, the Director of the National Institute of Health, and the Director of
the National Cancer Institute did not ask to join the case as interested par-
ties or intervenors (however, professional staff from HEW did testify in sup-
port of EPA's position). Thus from within government, EPA alone represented
the interest of the general public.
The adequacy of representation on behalf of the parties who bear the
present costs depends on how well financed were those representing the general
public. The Environmental Defense Fund was poorly financed, and this again
raises the issue of whether some means should be instituted for at least par-
tially compensating parties who successfully initiate or intervene on the side
of those who bear the present (or future) costs. There is precedent for com-
pensation within the legal process. In a class action brought by a few people
(or even a single individual) on behalf of other members of an affected class,
the attorneys representing the plaintiffs can expect to receive sizable fees
if they are successful. Pesticide cancellation and suspension cases, how-
ever, have not yet been brought as class action suits because there are no
specific damages or awards from which a fee could be paid.
To summarize, in this particular case we observe that once the proceed-
ing was instituted by the Environmental Defense Fund, EPA clearly represented
the public interest, and its side of the case appears to have been well fi-
nanced, well prepared, and well presented. _3/
2. Are the parties who obtain the present benefits represented? Signi-
ficantly, the Secretary of Agriculture filed to intervene in opposition to
the suspension order, kj along with special interest groups that included the
manufacturers, several dozen distributors of products containing chlordane
112
-------
TABLE 5
Criteria for Evaluating Decision Processes
*
Embodying Low Probability, High Consequence Risks
1. Are parties who incur the present costs adequately represented?
2. Are the parties who obtain the present benefits represented?
3. Are interests of future generations suitably considered?
4. Are the distributional aspects considered systematically?
5. Are the social decision-makers unbiased and open?
6. Is the development of knowledge concerning the effect of possible
alternative decisions adequate?
a. Do views that appear to come from reasonable sources receive
— or appear to receive — "equal time and treatment"?
b. Are all significant differences considering the quality of
knowledge included?
c. Are the gaps in knowledge identified?
d. Are procedures for collecting knowledge of the area efficient?
e. Are the plausible consequences spelled out for the worst contingency?
f. Are alternatives suitably considered?
7. Is information given to the individuals who are at risk?
8. Can a decision change if significant new information is obtained?
*
Source: Part I, Theoretical Considerations, Chapter 4, pp. 31-35.
113
-------
or heptachlor, and farm groups whose members used chlordane and heptachlor.
Thus all parties directly benefiting from the sale and use of the pesticide
were strongly represented, as was to be expected.
3. Are interests of future generations suitably considered? Decisions
concerning pesticides appear to involve the interests of future generations
only marginally, because the persistence or effective half-life of pesticide
residues is rarely over 3-5 years. For chlordane and heptachlor it was
estimated that if all usage ceased, then within three years the amount of
residue in foodstuffs would decline to an almost imperceptible trace. Follow-
up studies on DDT have shown that the amount of residues in the environment
declined fairly rapidly once usage was terminated. Thus persons affected by
residues from pesticides applied this year will be the present generation
plus, perhaps, children born during the next three or four years.
4. Are the distributional aspects considered systematically? Parties
opposed to the suspension of chlordane made a major point over the potential
impact on farmers who rely on chlordane/heptachlor products to control corn
cutworm (the "horizontal equity issue" discussed in Chapter 8). The argu-
ments were heavily weighed by the judge, who stated in his recommended deci-
sion that he "agonized over the impact which suspension would have."
Beyond the immediate impact on certain corn farmers, EPA introduced
evidence to show that suspension would have relatively little effect on total
output or the price of corn, since (a) chlordane products were used for in-
surance purposes on only 3% of the total corn crop, and (b) corn not pre-
treated with chlordane could be "scouted" at regular intervals after planting
and then treated with other pesticides if cutworm attacks were discovered.
Some corn is distributed to consumers through retail grocery outlets, and a
reduction in output might have some small effect on the retail price of corn
to consumers. Not all corn goes directly to consumers. A considerable amount
of corn grown in the United States is used to feed cattle and poultry. The
effect of a small reduction in output on the price of feed would be relatively
slight since soybeans and other crops are also used for feed. Beyond the
discussion about adjustments faced by corn farmers and price effects, there
was little systematic consideration of distributional effects. Similar
statements could be made about other crops for which chlordane or heptachlor
is used to control some specific pest (e.g., Hawaiian pineapple). To the
extent that price effects from suspension are in fact minimal, one might be
excused from such a discussion on the ground that it is not particularly
material to the proceeding. Even if the price effects had been significant,
one is left with the uneasy feeling that nothing more would have been sub-
mitted as evidence about the incidence of price changes in different income
groups within society.
The legal tradition is to worry about considerations of equity, but the
formal cost-benefit model ignores distributional effects and equity consider-
ations. This is an important difference between the legal and optimization
paradigms. Except for studies on the regressivity of taxes, the tradition in
many areas of economic analysis is to ignore distributional effects. In
Chapter 8 we suggested that it might be helpful to require one or both of the
two opposing sides to submit a formal cost-benefit analysis. Even if this
114
-------
were done, however, the distributional impacts of cancellation or suspension
might not be subjected to systematic consideration.
To summarize, distributional effects were considered, but not in any
systematic way. Moreover, even the most rigorous cost-benefit analysis will
ignore distributional effects. Therefore, in order to assure that distribu-
tional impacts are systematically taken into account, it might be helpful to
require both sides to submit a formal "distributional impact statement". This
impact statement would complement a cost-benefit analysis by indicating how
the estimated costs and benefits used in the analysis are distributed. Dis-
tributional impact statements are already routinely required, for example,
in studies prepared for the Congressional Office of Technology Assessment.
5. Are the social decision-makers unbiased and open? In the chlordane/
heptachlor case, and in any other pesticide case, the procedure is well estab-
lished. First, the case is heard by an Administrative Law Judge, who holds
an executive appointment and is usually a well-qualified professional whose
sole duty is to preside over administrative law proceedings. To prevent
judges from exercising open bias, strict rules exist regarding the disqualifi-
cation of judges. Should a judge with a known vested interest not disqualify
himself, the case can, on appeal, be overturned on this issue. There is thus
good assurance that this level of decision-making is reasonably free of bias.
However, although the Administrative Law Judge is doubtless quite influential
in shaping and determining the ultimate result, he is not the final decision-
maker. The Administrative Law Judge renders a lengthy, detailed opinion, con-
taining both the recommended decision and the major facts and reasoning which
underlie it.
The Administrative Law Judge's recommended decision is submitted to the
Administrator of EPA, who has responsibility for rendering a "final" decision.
To emphasize that it is the Administrator, and not the Administrative Law
Judge who makes the final decision, we point out that in the chlordane/
heptachlor case the Administrator did not accept the recommendation of the
Administrative Law Judge (who recommended against suspension) and instead
chose to suspend chlordane and heptachlor for most uses. The Administrator
carefully documented his reasons for not accepting the Administrative Law
Judge's recommendation.
The Administrator is a presidential appointee, like the heads of all
other government agencies. The appointment of this key individual will,
therefore, be subject to considerable political bargaining. Also, the Admin-
istratrator has numerous duties and responsibilities which extend far beyond
presiding over adversary proceedings. Because of his continued contact with
both environmental groups and industry, the Administrator will be subject to
many pressures, but will also have as much information and perspective on
environmental problems (especially policy issues) as anyone else in the coun-
try. For all these reasons, the Administrator can be expected to give a
carefully reasoned and balanced, if certainly not disinterested, decision on
all pesticide cases. This is probably the best that can be hoped for under
the circumstances. The explicitly political nature — and attendant public
responsibility — of the decision making process can be regarded as one of the
strong points of the existing institutional structure.
115
-------
Beyond the Administrator's decision, either party may appeal to the
courts. Federal judges are less subject to short-term political influence
than a political appointee in the executive branch of the government, but in
the longer run the courts have been sensitive to the currents of public opin-
ion. Federal judges will, however, be considerably less informed about envi-
ronmental problems than EPA's Administrator. Thus, to the extent that pesti-
cide cases are appealed to the federal courts, the decision makers will be
freer of bias and more open. This multi-tier review process helps greatly to
free the system of undue personal bias.
6. Is the development of knowledge concerning the effect of possible
alternative decisions adequate?
a. Do views that appear to come from reasonable sources receive
— or appear to receive — "equal time and treatment"? In the chlordane/
heptachlor case a large number of firms and organizations asked to join the
case in opposition to the proposed suspension order. The process is thus
open, but this openness also has a potential for great proliferation and du-
plication of effort. Looking beyond the chlordane/heptachlor case, the more
important point to emerge is that the legal process has the facility for con-
solidating large numbers of intervenors, and it can reasonably be expected to
use this facility in a way that preserves openness while promoting efficiency.
b. Are all significant differences considering the quality of
knowledge included? Cross-examination by knowledgeable persons is an effec-
tive and valuable means of bringing to light significant differences in the
quality of information submitted as evidence. In regard to this particular
criterion the adversary process appears to have advantages over, say, congres-
sional hearings. It would appear, nevertheless, that the adversary process
could be improved. In Chapter 9 we discussed the possibility of requiring
that all scientific data and testimony be upgraded to a publishable level,
and that such evidence in fact be published and made subject to more wide-
spread peer review. This procedure would help bring out any significant
differences in the quality of knowledge.
Improved pre-trial discovery procedures (e.g., requiring that expert
witnesses submit prepared testimony containing professional opinions in ad-
vance of oral testimony) and the introduction of scientific advisors to the
court might also improve the extent to which differences in the quality of
knowledge are brought out.
c. Are the gaps in knowledge identified? The legal process used
in pesticide cases can generally be counted on to produce a great deal of
evidence. The chlordane/heptachlor suspension hearings were conducted under
a "tight" time limit of 90 days (a subsequent 10-day extension was granted),
yet the record contained over 10,000 pages of testimony and evidence. It
would be naive to assume that a large bulk of evidence necessarily means that
knowledge, and gaps in knowledge, are adequately revealed by the process.
"Adequacy" is a relative term; the quality of evidence is what matters, not
just the volume. How adequate knowledge is depends on what other knowledge
could be made available during the decision process (i.e. , at the time of the
hearing) but is in fact not brought out.
116
-------
This raises a further issue. The availability of knowledge could, at
any time, be increased by defining and undertaking significant new research
projects. Such a potential increase in knowledge would involve both large
expenditures and time lags, and is not intended to be covered by the effort
designed to bring out existing knowledge. Yet there exists no sharp dividing
line between efforts aimed at "bringing out existing" knowledge and efforts
aimed at "increasing" knowledge. Useful "existing" knowledge is often buried
in a mountain of useless information from which it must be retrieved by large
expenditures of time and effort. In this regard the bringing out of existing
information for legal purposes is very much like the search for new knowledge.
Conversely, new research projects are very much like information retrieval,
especially in the area of social impacts, since they do not create information
out of a void, but gather information that is in some way hidden or encoded
within society or nature.
In the context of the adversary model used for suspension and cancella-
tion hearings, we can examine the question from the two viewpoints of those
opposed to and those favoring cancellation. One particular facet of adequacy
is then the issue of how well financed each side is. Those opposed to cancel-
lation will usually be well financed since this position represents concen-
trated interests, and in general they can be expected to submit in evidence
most of the knowledge, experimental outcomes, or other arguments favoring
their position. On the other side of the issue, the specter of the free-
rider looms large. In other words, where many individuals benefit from an
indivisible joint outcome that must be supported by their pooled voluntary
contributions, each individual is motivated to take a free ride on the efforts
of all the others, thereby undermining the viability of the enterprise. As
noted in our earlier discussion of the first evaluation criterion, the legal
process for pesticide decisions, per se, has no way of assuring any funding
whatsoever for the public position. All funding comes from the political pro-
cess that accompanies and underlies annual budgetary decisions.
In the chlordane/heptachlor case the EPA appears to have done a good job
of collecting evidence and mobilizing highly knowledgeable witnesses, insofar
as the effects of chlordane/heptachlor taken in isolation^ are concerned.
The availability of knowledge favoring cancellation arose largely from the
government's monitoring effort and government sponsored research, and the
mobilization of that knowledge on this issue was due mostly to the effort by
EPA.
The most critical gap in knowledge — never adequately brought out or
emphasized by the decision process — pertains to the possibly harmful syner-
gistic effect of chlordane/heptachlor in interaction with other environmental
or biological contaminants. When such synergistic effects are present, risks
owing to individual agents are not simply additive, but can enhance each
other manyfold. In areas of risk other than those involving pesticides,
known harmful synergies occur, for example, between smoking and urban air
pollution, smoking and the use of birth control pills, or the intake of bar-
bituates and alcohol. The suspension hearings never brought out our profound
ignorance concerning possible carcinogenic risks of chlordane/heptachlor resi-
dues in conjunction with smoking, with urban air pollution, with chemical food
additives, with common drugs, or with other pesticide residues. Perhaps no
117
-------
such synergies exist. But if any deadly synergies should exist, the current
decision process can certainly fail to take them into account.
d. Are procedures for collecting knowledge of the area efficient?
The avoidance of serious overlap and duplication does not close the issue of
efficiency. The proceedings of the chlordane/heptachlor suspension case gen-
erated a substantial amount of material, and the volume of information that
will be generated during the full cancellation proceeding can be expected to
dwarf that of the suspension hearings. The information which is submitted is
carefully catalogued and accessible, in the best legal tradition. The ques-
tion is: Could the bulk of the most important information be brought together
in a more efficient way? In this regard, we have the feeling that there might
be room for improvement. For example, the law stipulates that the Administra-
tor must consider benefits — i.e., cost-risk-benefit analysis is specified as
the basic decision framework — yet neither side in the suspension hearings
actually prepared and submitted a cost-benefit analysis. Voluminous informa-
tion on costs and benefits was submitted, in what might be described as piece-
meal fashion, but there was no single vehicle for bringing together and con-
solidating all this information. No formal cost-benefit model was submitted
in advance, although one most certainly could have been specified. Similarly,
as we pointed out previously, there was no attempt at formal risk analysis.
What might be helpful, therefore, would be to require one or both parties to
submit a formal cost-benefit analysis, complete with estimates of all future
benefits and costs, discount factors, breakeven analysis, etc., in much the
same way that environmental impact statements are now required in many areas.
The extent to which both sides either agreed or differed with the analysis
contained in a formal model might be of great help in narrowing and focusing
the issues. Formal specification of risk-benefit models, especially if done
within the adversary setting, might in time also help broaden public famili-
arity with and acceptance of the optimization paradigm.
Another issue relating to the identification of paps in knowledge con-
cerns whether a sufficient number of the "right" questions were raised during
the hearing. As the discussion about possible science advisors to the court
pointed out, the existing adversary process relies almost totally on opposing
counsel to ask the right questions, but in highly technical matters this ap-
proach may be deficient. Given the preponderance of lawyers present at such
hearings, there is good assurance that all pertinent legal questions will be
raised. However, counsel for both sides may fail to ask some highly relevant
scientific questions, particularly when the questions and issues are of a
technical nature. Scientific advisors to the court, coupled with the ability
of the judge to ask questions pertaining to issues not adequately covered or
not otherwise raised, would provide additional assurance that adequate know-
ledge will be divulged by the process.
To summarize, the knowledge brought out in the chlordane/heptachlor case
appears to have been generally adequate, but this may have been an exceptional
situation. It resulted from the adequate financing provided by EPA, and the
cooperation of several witnesses from other government agencies. When affec-
ted parties on one side of an issue are virtually all free-riders, then the
legal process — as presently constituted — has few built in assurances that
adequate knowledge will be forthcoming.
118
-------
e. Are the plausible consequences spelled out for the worst con-
tingency? In the chlordane/heptachlor suspension hearing the plausible con-
sequences were not clearly spelled out for the worst contingency. As noted
-4
previously, one witness stated that a probability of 10 in a population of
Q
2 x 10 would result in 20,000 deaths, but that was not put forth as either
his best estimate of the risk or as the worst contingency.
As with all low-probability, high-consequence outcomes, there are a great
many unknowns concerning the risks associated with pesticide residues. It is
not known, for instance, whether chlordane/heptachlor residues are more likely
to induce some types of cancer more than other types; nor do we know how many
years of continued exposure will, on the average, result in a known number of
cases. As indicated before, we know nothing about synergistic effects.
Still, it would seem helpful to have some concept of the worst possible con-
sequences spelled out. In some instances the spelling out of the worst pos-
sible consequences (to be compared with the potential benefits from continued
usage) might be more helpful to those advocating registration, and in other
instances it might be more helpful to those advocating suspension and cancel-
lation. We do not know which side would have been helped in the chlordane/
heptachlor case, and this discussion is not intended to fault either side in
the preparation and presentation of its case. If there is any failing, it is
that the process itself does not require this type of information to be pre-
pared and submitted.
f. Are alternatives suitably considered? Does the process expli-
citly recognize the uncertainty and lack of knowledge about pertinent alter-
natives or possibilities? In a pesticide suspension or cancellation proceed-
ing, such as the chlordane/heptachlor case, the range of alternatives
considered is narrow. The principal alternatives to non-use of the pesticide
at issue currently are:
* The use of other pesticides or other pest control tech-
niques against those pests which are known to be con-
trolled by the pesticide whose registration is being
challenged.
* Growing other crops which are not susceptible to the
pest controlled by the pesticide whose registration
is being challenged.
As indicated in Chapter 7, one of the principal agricultural uses of
chlordane, especially in low-lying areas with poor drainage, is to control
the corn cutworm, which has a devastating effect on a corn field if uncon-
trolled. The alternatives are to use other means of controlling the corn
cutworm, or to grow some other crop that will not be affected by the corn
cutworm, such as soybeans. During the chlordane/heptachlor suspension case
extensive evidence was presented on both alternatives. However, both the
recommended decision by the Administrative Law Judge and the final decision
by the Administrator focused their attention almost exclusively on other means
that are available to control the cutworm (and that would permit those farmers
using chlordane/heptachlor to continue growing corn) , and gave scant attention
119
-------
to the possibility of growing other remunerative crops on the same land. On
the basis of this one case study, we do not know whether this result is a
peculiarity of the chlordane/heptachlor case, or whether there is a subtle
bias in the law which gives heavy weight to those alternatives which would
preserve both the "physical substance" as well as the economic substance of
a farmer's situation. If such a bias exists, then it might be useful: (a)
to have it brought out into the open and discussed; and (b) to develop formal
checklists or models for use in such cases to help prevent important factors
(such as good economic alternatives) from being systematically omitted. In
light of the extensive evidence which the judge must sift in such cases, it is
entirely understandable that something like alternative uses of the land might
escape attention in the final decision. However, such a relatively narrow
focus may also be the result of a systematic bias.
As regards the particular alternatives discussed above, the adversary
system is capable of developing the necessary detail and bringing many to the
attention of the decision makers. This was in fact done in the chlordane/
heptachlor case. There are instances of broader policy alternatives available
to society but they fall outside the scope of the legal process specified in
the applicable legislation. Such alternatives will therefore be irrelevant
and immaterial to the legal proceeding. The application of chlordane to con-
trol the corn cutworm provides a good example of the possibility of applying
broader policy alternatives. Corn growers apply chlordane in the early
spring, before the crop is planted, and before it is known whether cutworms
will be a problem that year. In other words, chlordane is applied as "insur-
ance" against cutworm infestation. As occurs with all "insurance" policies
for casualty-type events, only some will (or would) suffer from the casualty
event which is being insured against. Just as most home-owners who buy in-
surance will not suffer from fires, so most corn farmers who use chlordane
would not actually suffer from cutworm infestation. The fact that the use of
chlordane is for insurance purposes, and not to control a known infestation,
gives rise to a broader social alternative. Namely, create and sell to corn
farmers an actual insurance policy that would cover cutworm infestation. It
is entirely possible that the cost of such insurance would not be dispropor-
tionately greater than the cost of buying and applying chlordane, and it could
well turn out to be quite a bit less. This would depend upon the extent of
damage that would actually be done by cutworms (assuming that no chlordane
were used). If the cost turned out to be lower, the advantage is obvious; if
higher, some extra cost could still be advantageous from the social point of
view in order to reduce the environmental impact of chlordane. 5J However,
as indicated above, any consideration of such alternatives is now beyond the
purview of an administrative law proceeding.
The narrow conception of alternatives under the current decision-making
process makes it all but impossible for some of the key dynamic effects of
pesticide decisions to be raised in a legal proceeding in a relevant and
material manner. These dynamic effects, discussed earlier, include: (1)
effects of research and development aimed at new pesticides; (2) effects on
the entry of new firms into the industry; (3) effects on research and develop-
ment aimed at new non-pesticide pest control alternatives; and (4) effects on
social adjustments modifying the entire interface between food production,
resource use, and the total ecology. Raising issues connected with each of
120
-------
these effects implies consideration of the use or non-use of a given pesticide
in the context of an extremely broad range of social alternatives. Moreover,
the effects in question cannot be properly tied down to a decision concerning
any single pesticide. They hinge much more on the entire economic and social
climate created by a series of legally independent but substantially inter-
related pesticide decisions. The current decision-making process effectively
excludes such issues from consideration, yet over a horizon of just a few
years these issues may have far more social importance than those laboriously
considered.
7. Is information given to the individuals who are at risk? The legal
process followed in pesticide cancellation and suspension proceedings estab-
lishes an extensive public record, which means that information brought out
by the process is in principle available to individuals who are at risk. Yet
this record is not published and circulated in printed form like the record
of congressional hearings. To the extent that individuals at risk are repre-
sented at the hearings, as by the Environmental Defense Fund, they may receive
information brought out during the hearings through publications issued by
such organizations. Also, while the trial process itself typically receives
little publicity, the final decision rendered by the Administrator tends to be
"news", and this causes it to be picked up and disseminated by the general
news media. Thus the process would appear to conform reasonably well to this
criterion.
8. Can a decision change if significant new information is obtained?
The registration of a previously registered pesticide can clearly be cancel-
led. Under EPA rules and regulations the process for reconsidering and rein-
stating the registration of a pesticide is generally similar to that for
initial registration. Namely, a manufacturer, potential user, or other inter-
ested party (e.g., the Department of Agriculture) initiates the process by
petitioning the Administrator for reconsideration. The basis for such recon-
sideration would be substantial new evidence as to the benefits or risks of
the pesticide. Such evidence would be submitted with the petition, and re-
viewed by EPA. If the Administrator should decide that reconsideration is
warranted, a full administrative law proceeding, open to the public, is then
undertaken. This hearing can be as extensive as a full cancellation proceed-
ing. Following the hearing the Administrative Law Judge submits a recom-
mended decision to the Administrator, who issues a final order. Decisions of
the Administrator can be appealed to the Federal District Court.
Summary and Conclusion
This case study has focused on the process by which social choices are
made for one specific type of low-probability, high-consequence risk. The
study has dealt with the decision to continue or suspend the registration of
a pesticide whose residues are known to enter the human biomass and perhaps,
as shown by limited evidence, exert carcinogenic effects. Specifics of the
chlordane/heptachlor suspension case have been used as a means of gaining in-
sight into the decision process and generalizing about it. At the same time
we have discussed the decision process from a broader social point of view in
order to put the EPA suspension-cancellation process into better perspective.
The efficacy of a social decision process cannot be measured on an absolute
121
-------
scale; it must, rather, be judged in relation to other decision processes
that are available to us.
Cost-risk-benefit analysis, as used in the chlordane/heptachlor case,
was described and appraised in Chapter 8. In Chapter 9 the label "optimiza-
tion paradigm" was applied to the principle underlying the cost-benefit ap-
proach, and it was contrasted to the "legal paradigm", which constitutes the
basic framework of the current decision process. Both approaches were seen
to have shortcomings.
The optimization approach attempts to determine the nominally best social
choice for a given set of facts and circumstances. It has not, however, gone
through the long and careful development of the legal approach. Its methods
tend to be ad hoc rather than standardized; it is subject to few agreed-upon
procedures or rules; its assumptions and the limits it places upon the scope
of its models often seem arbitrary; and its conclusions are frequently colored
by the personal biases of the individuals using the tools and techniques. For
these reasons the optimization approach has limited acceptance or authority
in social decisions.
The legal approach has deep roots in the lengthy development of civil
and criminal law. The primary purpose of the legal process is to resolve
conflicts according to the.principle of justice under the law, not to seek
some kind of "optimum" or "best" social decision. Accordingly, the legal
process was developed to adjudicate and settle issues relating to events that
have already taken place. It seems fair to say that the legal process was
never intended to create a forum for gathering and weighing scientific infor-
mation related to future outcomes, especially when those future events are
subject to a high degree of uncertainty. Thus Congress has now mandated the
generalization of the legal-adversary model to a situation for which it was
neither designed nor intended. A major strength of the adversary model, how-
ever, is that the legal process enjoys broad social acceptance, regardless of
whether it arrives at a "correct" or "optimal" decision in every instance.
Since Congress has had no ready alternatives at its disposal, this broad ac-
ceptance is doubtless a major reason why it opted for the legal process.
In this chapter we have evaluated the current decision process — which
represents a marriage of the optimization and the legal paradigms in their
existing mode — against the eight criteria developed in Part I of the study.
Generally speaking, as evidenced by the chlordane/heptachlor suspension case,
the current decision process is seen to create a reasonable and acceptable
way of dealing with the cancellation or suspension of individual pesticides,
taken one by one, and considered against a narrow range of alternatives.
The current procedure can, however, offer no assurance against harmful
synergistic effects of environmental contaminants in combination, nor can it
guarantee that its outcomes will be reasonable and acceptable when considered
against a full range of social options.
EPA is at the leading edge of societal change. It is inevitable that
many issues which EPA must face will involve low-probability, high-consequence
occurrences. Thus EPA is now and will continue to be the forum for resolving
122
-------
issues involving major areas of scientific and technical uncertainty, coupled
with deep-seated conflict. Processes for dealing with this type of issue are
the suject of concern and study by government agencies other than EPA. As
these issues become of increasing concern to an ever growing segment of the
population, social acceptability of the decision process is of paramount
importance.
This study points to three recommendations. First, EPA should expand
its research dealing with the decision process itself, particularly in the
areas of synergistic effects and broad social options where the decision pro-
cess is now weakest. Second, EPA should monitor, on a continuing basis, modi-
fications and improvements in the process which may evolve elsewhere: in the
federal government, in other public or private decision making bodies of this
country, and in decision making agencies or overseas organizations that are
working under a broad range of social and economic conditions. Third, EPA
should establish a capability for continuous self-evaluation of its major
programs that embody decision processes, focusing on its own success so far
as compared with other major U.S. regulatory agencies, in maintaining a pos-
ture of independence in relation to the key interest groups subject to its
regulatory action.
Notes to Chapter 10
1. The term "inquisitorial" is not used in any derogatory or pejorative
sense. It is merely a means of distinguishing one type of hearing from
an adversary-type hearing.
2. The risks and costs may sometimes be localized, as with the PCB incident
in Michigan. But in view of the widespread network through which most
crops are distributed, the situation represented by the chlordane/
heptachlor case would appear to be more typical.
3. One should nevertheless be careful not to extrapolate too far beyond one
case. In the past, political forces have managed to keep consumer agen-
cies chronically underfunded. Also, among the ranks of regulatory agen-
cies, EPA is still quite young. Many regulatory agencies, as they
mature, come increasingly under the influence of the industries they have
been created to regulate. Such influence is acquired when the regulated
industry retains, at fees substantially higher than those paid by govern-
ment, lawyers, economists, engineers and scientists who appear to be most
knowledgeable and capable in the areas of regulation concerned. The
"public interest" provides few jobs in the environmental area at places
like the Environmental Defense Fund and Natural Resources Defense Council
— and at comparatively low salaries. The EPA, of course, regulates many
more industries than most regulatory agencies, and it may therefore re-
sist capture by the affected interests longer and more successfully than
other regulatory agencies . While EPA did a creditable job in this case,
it is too early to advocate that EPA be the first and last line of public
defense in environmental matters.
123
-------
A. The Department of Agriculture intervened because it has used chlordane/
heptachlor products in some of its quarantine programs.
5. Secretary of Agriculture Robert Bergland is reported to favor such an
approach. According to the New York Times, Secretary Bergland plans to
develop an "all-risk, all crops" insurance program to protect producers
against disasters of the kind that had occurred in many parts of the
country in the past year. According to Secretary Bergland:
"The disaster provisions of our present laws are in and
of themselves a disaster. . .
"The Secretary said that he had been advised by private
insurance companies that the risks of insuring crops were
beyond their financial resources and, therefore, coverage
was a federal responsibility."
(New York Times, February 1, 1977, p. 17).
124
-------
References for Part II
Arrow, Kenneth A. "Notes on Evaluation of Publicly Manageable Risks to
Life." First draft. February 22, 1976.
Arrow, Kenneth A., and R. C. Lind. "Uncertainty and the Evaluation of Public
Investment Decisions," American Economic Review, 60 (June, 1970),
pp. 364-378.
Carter, Luther J. "Pest Control: NAS Panel Warns of Possible Technological
Breakdown," Science, 191 (February 27, 1976), pp. 836-837.
Columbia Broadcasting System. CBS Reports: "The Politics of Cancer," trans-
script of a 30-minute broadcast program, June 22, 1976, 10:00 - 11:00
PM, EOT, with CBS News Correspondent Lesley Stahl. c 1976.
Council for Agricultural Science and Technology. The Environmental Protec-
tion Agency's Nine "Principles" of Carcinogenicity. Ames, Iowa, Council
for Agricultural Science and Technology, Department of Agronomy, Iowa
State University, January 19, 1976.
Council for Agricultural Science and Technology. Pesticide Report to the
Nation, proceedings of a symposium, Chicago, Illinois, 1973. [Ames,
Iowa, Department of Agronomy, Iowa State University] March, 1974.
Debreu, Gerard. Theory of Value: On Axiomatic Analysis of Economic Equi-
librium. New York: Wiley, 1959.
Federal Working Group on Pest Management, Monitoring Panel. Catalog of
Federal Pesticide Monitoring Activities in Effect July 1973. March,
1975. (Available from National Technical Information Service, Spring-
field, Va. 22151.) (FWGPM-75-1).
Halter, A. N,, and G. W. Dean. Decision Under Uncertainty. Dallas: South-
western Publishing Company, 1971.
Heilbroner, Robert Louis. The Economic Problem. 3rd ed. Englewood Cliffs,
N. J.: Prentice-Hall, 1972.
Hirschleifer, Jack. "Investment Decisions Under Uncertainty: Applications
of the State Preference Approach," Quarterly Journal of Economics, 80
(May, 1966), pp. 252-77.
125
-------
Hoel, David G. , and others. "Estimation of Risks of Irreversible Delayed
Toxicity," Journal of Toxicology and Environmental Health, 1 (1975),
pp. 133-151.
Kornai, Janos. Anti-Equilibrium; On Economic Systems Theory and the Tasks
of Research. New York: American-Elsevier, 1972.
Kuhn, Thomas S. The Structure ofScientific Revolutions^ 2d ed. enlarged.
Chicago: University of Chicago Press, 1970 (reprinted from Interna-
tional Encyclopedia of Unified Science, vol. 2, no. 2).
Lancaster, K., and R. Lipsey. "The General Theory of the Second Best,"
Review of Economic Studies, XXIV(1), No. 63, pp. 1-18.
Linnerooth, Joanne. "The Evaluation of Life-Saving: A Survey." Laxenburg,
Austria: International Institute for Applied Systems Analysis. July,
1975. (Research Report -75-21).
Majone, Giandomenico. "Choice Among Policy Instruments for Pollution Con-
trol," Policy_Analy_si£ 2 (Fall, 1976), pp. 589-614.
McKean, Roland N. The Regulation of Chemicals and the Production of Informa-
tion . Working Paper No. 13. St. Louis: Center for the Study of
American Business, Washington University, June, 1976.
Mantel, Nathan, and Marvin A. Schneiderman. "Estimating 'Safe' Levels, a
Hazardous Undertaking," Cancer Research, 35 (June, 1975), pp. 1379-1386.
Mantel, Nathan, and W. Ray Bryan. "'Safety1 Testing of Carcinogenic Agents,"
Journal of the National Cancer Institute, 27 (August, 1961), pp. 455-470,
Morse, Chandler. "Depletion, Exhaustibility, and Conservation," Economics
of the Minerals Industries. In Preparation.
National Research Council. Committee on Principles of Decison Making for
Regulating Chemicals in the Environment. Decision Making for Regulating
Chemicals in the Environment. Washington, D.C.: National Academy of
Sciences, 1975.
National Science Foundation, Division of Social Sciences, Economics Program.
Conference on Econometrics and Mathematical Economics, Workshop on
Energy-Related General Research in Micro-Economics ]_ at _/ Conference
On Natural Resources, Palo Alto, California, May 9-11, 1975. Washing-
ton, D.C.: National Bureau of Economic Research, 1975.
"Pests and Pesticides: Deciding Between Them," News Report, National Academy
of Sciences. . ., XXVI (February, 1976), pp. 1-2.
Rowe, W. D. "An 'Anatomy' of Risk," Washington, D.C.: Environmental Pro-
tection Agency, March, 1975. Limited ed. draft.
126
-------
Schneiderman, Marvin A., Nathan Mantel, and Charles C. Brown. "From Mouse to
Man — Or How to Get from the Laboratory to Park Avenue and 59th
Street," Annals of the New York Academy of Sciences, 246 (January 31,
1975), pp. 237-248.
Specter, Phillip L. "Regulation of Pesticides by the Environmental Protection
Agency," Ecology Law Quarterly, 5:233 (1976), pp. 233-263.
Spencer, Donald A. The National Pesticide Monitoring Program, An Overview of
the First Ten Years of the Program's Operation. Washington, D.C.:
National Agricultural Chemicals Assn. 1974.
Stael von Holstein, Carl-Axel S. "A Tutorial in Decision Analysis," paper
presented (in slightly different form) at the Third Research Conference
on Subjective Probability, Utility, and Decision Making, London, Septem-
ber 7-9, 1971. Menlo Park: Stanford Research Institute, July, 1973.
U.S. Congress. House. An Act to Amend the Federal Insecticide, Fungicide,.
and Rodenticide Act of 1972. ("Federal Environmental Pesticide Control
Act"), Pub. L. 94-140, 94th Cong., H.R. 8841, November 28, 1975.
U.S. Congress. House. A Bill to Regulate Commerce and Protect Health and
the Environment by Requiring Testing and Necessary Restrictions on
Certain Chemical Substances and Mixtures, and for Other Purposes, ("Toxic
Substances Control Act"), H.R. 14032, 94th Cong., 2d Sess., 1976.
U.S. Congress. House. Committee on Interstate and Foreign Commerce. Toxic
Substances Control Act. H.R. Rept. 1341 To Accompany H.R. 14032, 94th
Cong., 2d Sess., 1976.
U.S. Environmental Protection Agency. "Agency Approach to Cancer Policy,"
internal memorandum from Russel E. Train, Administrator, to Assistant
Administrators, Regional Administrators, Office Directors, October 10,
1975.
U.S. Environmental Protection Agency. Analysis of Collective Social Choices
with Technological Risk and Uncertainty — Cost Proposal, submitted by
the National Planning Association, Washington, D.C.: June, 1974.
U.S. Environmental Protection Agency. Analysis of Collective Social Choices
with Technological Risk and Uncertainty — Technical Proposal, submitted
by the National Planning Association, Washington, D.C.: June, 1974.
U.S. Environmental Protection Agency. DDT; A Review of Scientific and
Economic Aspects of the Decision to Ban Its Use as a Pesticide. Wash-
ington, D.C.: Environmental Protection Agency, 1975. (EPA-540/1-75-022)
(For sale by National Technical Information Service, Springfield, Va.
22151.
U.S. Environmental Protection Agency. "Draft Report on Other Hazard Evalua-
tion Systems," internal report from Don Marlow, Program Manager, Hazard
127
-------
Evaluation System, to Director, Technical Services Division, November 7,
1975.
U.S. Environmental Protection Agency. Environmental Facts — DDT and the
Environment. Washington, D.C.: Environmental Protection Agency, July,
1974.
U.S. Environmental Protection Agency. "Interim Procedures & Guidelines for
Health Risk and Economic Impact Assessments of Suspected Carcinogens,"
internal memorandum from Russell E. Train, Administrator, to Assistant
Administrators, Regional Administrators, and Office Directors. Washing-
ton, D.C.: May 13, 1976.
U.S. Environmental Protection Agency. Internal document prepared by Don
Marlow on "A Pesticide Hazard Evaluation System," May, 1974.
U.S. Environmental Protection Agency. Internal information memorandum from
Deputy Assistant Administrator, Office of Pesticide Programs, on "Pesti-
cide Cancellation and Suspension Procedures," December 9, 1975.
U.S. Environmental Protection Agency. Internal memorandum from Arnold
Aspelin on "Thinking on Risk/Benefit Criteria and Decision Making,"
February 2, 1975.
U.S. Environmental Protection Agency. Internal memorandum from Don Marlow,
Project Manager, Hazard Evaluation System on "OPP and OE Data Sub-
systems," May 19, 1976.
U.S. Environmental Protection Agency. Internal memorandum from F. W. Talcott
re "Review of Abt Associates Crop Insurance Report." June 28, 1974.
U.S. Environmental Protection Agency. Internal memorandum from F. W. Talcott
re "Strategy Implications of Abt Crop Insurance Study." July 1974.
U.S. Environmental Protection Agency. Internal report by Arnold Aspelin and
Fred Arnold on "Economic and Social Aspects of Pesticide Regulation;
Current Efforts and Challenges, EPA Economists Briefing." May 22, 1975.
U.S. Environmental Protection Agency. "Pesticide Program: Guidelines for
Registering Pesticides in the United States," Federal Register, v. 40,
no. 123, Part II, June 25, 1975, pp. 26802-26928.
U.S. Environmental Protection Agency. "Pesticide Programs: Registration,
Reregistration and Classification Procedures," Federal Register, v. 40,
no. 129, Part II, July 3, 1975, pp. 28242-28286.
U.S. Environmental Protection Agency, Office of Pesticide Programs. Alterna-
tive Futures for Environmental Policy Planning: 1975-2000. A report
prepared by Stanford Research Institute, Center for the Study of Social
Policy. Washington, D.C.: Government Printing Office, October, 1975;
Springfield, Va.: National Technical Information Service, October, 1975.
(EPA-540/9-75-027).
128
-------
U.S. Environmental Protection Agency, Office of Pesticide Programs. _A
Benefit-Cost System for Chemical Pesticides, a report prepared by CONSAD
Research Corp., Washington, D.C.: Environmental Protection Agency, June,
1975. (EPA-540/9-76-001).
U.S. Environmental Protection Agency, Office of Pesticide Programs. Incre-
mental Cost Impacts of the 1972 Federal Insecticide, Fungicide and
Rodenticide Act as Amended. Washington, D.C.: Environmental Protection
Agency, 1976. (EPA-540/9-76-002).
U.S. Environmental Protection Agency, Office of Pesticide Programs. "The
National Pesticide Monitoring Plan — Draft," September 11, 1975.
U.S. Environmental Protection Agency, Office of Pesticide Programs. Strategy^
of The Environmental Protection Agency For Controlling the Adverse
Effects of Pesticides. Washington, D.C.: Environmental Protection
Agency, May, 1974.
U.S. Environmental Protection Agency, Office of Pesticide Programs. A Study
to Develop Alternative Conceptual Approaches for a Pesticide Hazard
Evaluation System, final report prepared by Arthur D. Little, Inc.,
Cambridge, Mass., April, 1976.
U.S. Environmental Protection Agency, Office of Pesticide Programs, Criteria
and Evaluation Division, Economic Analysis Branch. Internal memorandum
on "EPA Pesticicide Cancellation/Suspensions: A Survey of Economic
Impacts on Users," September, 1975.
U.S. Environmental Protection Agency, Office of Pesticide Programs, Criteria
and Evaluation Division, Economic Analysis Branch. Internal memorandum
on "Framework for Risk/Benefit Analysis of Pesticides in Problem Use
Patterns," February, 1976.
U.S. Environmental Protection Agency, Office of Pesticide Programs, Ecological
Monitoring Branch. Pesticide Monitoring Quarterly Report, nos. 1, 2 and
3 (July-December, 1975 and January-March, 1976).
U.S. Environmental Protection Agency, Suspension Proceedings Before the
Administrator, in re: Velsicol Chemical Corporation et al., Registrants,
FIFRA Docket No. 384, Clarification of Order of December 24, 1975, by
Michael K. Glenn, Chief Judicial Officer, January 19, 1976.
U.S. Environmental Protection Agency, Suspension Proceedings Before the
Administrator, in re: Velsicol Chemical Corporation et al., Registrants,
(Expedited Hearing on Heptachlor-Chlordane). FIFRA Docket No. 384,
Decision of the Administrator on the Suspension of Heptachlor-Chlordane,
by Russell E. Train, December 24, 1975.
U.S. Environmental Protection Agency, Suspension Proceedings Before the
Administrator, in re: Velsicol Chemical Corporation et al., Registrants,
Recommended Decision, by Herbert L. Perlman, Chief Administrative Law
Judge, December 12, 1975.
129
-------
U.S. Environmental Protection Agency, Suspension Proceedings Before the
Administrator, in re: Velsicol Chemical Corporation et al., Registrants,
Statement for Testimony at Chlordane/Heptachlor Suspension Hearing,
by Dr. Arnold L. Aspelin, Chief, Economic Analysis Branch, Criteria
and Evaluation Division, Office of Pesticide Programs, September, 1975.
Zeckhauser, Richard. "Procedures for Valuing Lives," Public Policy, 23
(Fall 1975), pp. 419-464. Also available as Reprint No. 104R in Harvard
University, School of Government Reprint Series.
130
-------
Part III
CASE STUDY OF NUCLEAR WASTE DISPOSAL
131
-------
11. Introduction
A degree of consensus exists among experts as well as the general public
that the rate of adaptation of new technologies and their use in human en-
deavors have increased during recent times. _!/ There also appears to be an
increase in public awareness of the presence of societal risks resulting from
man-made activities 2j as well as those derived from natural phenomena. 3J
This is a case study of societal risk management as practiced for nuclear
waste disposal. Seldom has an issue associated with technological risk and
uncertainty been more widely debated by the public than the subject of nuclear
waste disposal. Further, both the public and the scientific community have
generally assumed that risks associated with radioactive waste disposal can
either be readily overcome by existing technologies or that catastrophy to
mankind is inevitable. These radically different views are held not only by
the general public but also by nuclear scientists, engineers, economists, and
elected officials. The concern and differences of opinion on risks associated
with nuclear wastes are significant and well informed.
The Catastrophic View
The Committee of the Pugwash Conferences on Science and World Affairs
gave the following evaluation of man's current ability to store high-level
wastes in perpetuity:
No general solution for the isolation of long-lived
radioactive wastes from the biosphere, for the necessary
many thousands of years, is yet in hand; that is, despite
a wide variety of proposals, 'experts' still disagree on
whether any of them will suffice. . . It is impossible to
be complacent about expansion in the use of nuclear
power without having a solution in hand. A_/
A similar conclusion was reached by the Union of Concerned Scientists:
The matter of man-made vaults at or near the surface for
long term storage (500 years) is. . .uncertain. Their
maintenance depends upon the existence of social and
political institutions, the permanence of which cannot
be guaranteed over time.
The impression is inescapable, in view of the present
imprecise state of affairs, that no convincing statements
exist regarding the long term environmental impact attend-
133
-------
ing the storage and/or disposal of wastes from fuel
reprocessing. It is disturbing in this regard that
the pace of development of a suitable waste disposal
technology does not seem commensurate with the rate of
expansion of the nuclear power industry. . . Fission
technology requires that man issue guarantees on events
far into the future, and it is not clear in most cases
how this can be done. Institutional arrangements do
not exist and never have existed to guarantee the moni-
toring of or attendance upon storage facilities over a
millennium. In the range of a million years, serious
geological uncertainties arise and even the survival of
man may be doubtful. 5_/
Distinguished scientists have also been outspoken to the effect that
there is no assured means currently available to store high-level wastes "per-
manently." For instance, Hannes Alfven, 1970 Nobel laureate in physics, made
the following statement:
At present there does not seem to be any existing, realis-
tic project on how to deposit radioactive waste; but there
are a multitude of optimistic speculations on how to do so.
The problem is how to keep radioactive waste in storage
until it decays after hundreds or thousands of years. The
deposit must be absolutely reliable as the quantities of
poison are tremendous. It is very difficult to satisfy
these requirements for the simple reason that we have had
no practical experience with such a long term project.
Moreover, permanently guarded storage requires a society
with unprecedented stability (T)he production of
nuclear energy is necessarily associated with the produc-
tion of radioactive elements; and a very large production
of nuclear energy necessarily means the mass production of
radioactive poisons in quantities which are terrifying.
This is the basic reason for the opposition against the use
of atomic energy, which has become a worldwide controversy. 6J
These radioactive wastes remain toxic for a very long period of time —
in some cases for hundreds of thousands of years. This means that they must
never be permitted to escape containment. As pointed out by J. 0. Blomeke
et al. in a 1973 article, an inescapable commitment of the use of fission
reactors for power generation is the management of radioactive wastes wher-
ever they are placed. The management of nuclear wastes (from all types of
fission reactors) requires isolation and/or containment until the wastes,
through radioactive decay, become relatively stable and less hazardous. The
time required for this is many thousands of years. Thus, the safety of waste
management concepts must be evaluated in a geological time frame. _7/
134
-------
According to current projections, several hundred billion curies of high-
level radioactive wastes vill have accumulated by the beginning of the next
century. These wastes will contain over 1000 metric tons of actinides, the
highly toxic, long-lived radionuclides that pose the bulk of hazards after
several hundred years of storage. By the year 2000, 50,000 trillion cubic
meters of water would be required to dilute the high-level radioactivity with-
in standards set by the Radiation Concentration Guides. Some respected
nuclear scientists are concerned over the ultimate disposal of these wastes.
A. S. Kubo and D. J. Rose state:
"The use of mausolea. . .is a temporizing measure. All
countries currently using this concept intend to use some
other method of disposal when new technology is developed
or the wastes are more manageable." 8_/
"Storage in man-made vaults is generally considered to be
useful only for interim storage since surveillance and
controls are required." _9/
A summary of the concerns regarding nuclear waste risks was recently
stated by Philip Micklin.
A key unsolved problem of the nuclear age is the centuries-
long management of extremely hazardous high-level radio-
active wastes produced in reactors. Until recently, stor-
age of liquids in underground high integrity tanks has been
the prevalent repository mode for such materials. Tanks at
Hanford and Savannah River have leaked wastes into the en-
vironment and the AEC* admits this is not a viable long-term
management mode. . .
In spite of unsolved problems, the AEC is convinced that
long-term management of highly radioactive wastes poses no
environmental hazards. They contend they, or some successor
organization, will be able to provide the careful control of
these materials required into the distant future or until a
reliable permanent disposal system is found. On this assump-
tion, no threat is admitted in the expected rapid build-up of
wastes from power reactors over the next several decades.
The AEC may be right. But if wrong, the consequences, borne
primarily by future generations, will be catastrophic. As
management and engineering systems are devised by humans, no
matter how carefully designed they are inherently subject to
failure, particularly over the long term. Hence, it may be
prudent to consider stopping or at least slowing the head-
long expansion of nuclear power and devoting our attention
to developing less hazardous alternative energy sources until
Atomic Energy Commission
135
-------
more complete answers are provided to this as well as other prob-
lems of nuclear technology. 10/
Micklin's plea for more complete answers may be difficult to satisfy, if it is
indeed true that these problems are beyond analysis (at least conventional
risk analysis). These views have also been expressed by Alvin M. Weinberg,
former Director of the AEC's Oak Ridge National Laboratory. He identified the
question of whether to generate the highly toxic, long-lived wastes as "trans-
scientific", a ". . .question which can be adjudicated by a legal or political
process, rather than by scientific exchange among peers." ll/ Weinberg fur-
ther stated:
We nuclear people have made a Faustian bargain with
society. On the one hand, we offer — in the catalytic
nuclear burner — an inexhaustible source of energy. . .
But the price that we demand of society for this magical
energy source is both a vigilance and a longevity of our
social institutions that we are quite unaccustomed to. .
The society must then make the choice, and this is a
choice that we nuclear people cannot dictate. We can
only participate in making it. Is mankind prepared to
exert the eternal vigilance needed to ensure proper and
safe operation of its nuclear energy system? 12/
Allen Kneese addressed the problem from an economist's perspective:
It is my belief that benefit-cost analysis cannot answer
the most important policy questions associated with the
desirability of developing a large-scale, fission-based
economy. To expect it to do so is to ask it to bear a
burden it cannot sustain. This is because these questions
are of a deep ethical character. Benefit-cost analysis
certainly cannot solve such questions and may well obscure
them. 13/
Actual past experience in waste management compared to future plans and pro-
posals does not in general indicate that appropriate nuclear management tech-
niques have been developed or are being used. On the contrary, past ap7
praisals of waste management methods, conducted by the General Accounting
Office (GAO), suggest that serious management and/or technical problems exist.
In this regard, information from two GAO reports should be noted:
At the time of our review, Richland was faced with a
potentially serious situation with respect to the con-
dition of its existing tanks. The operating contractor
has estimated that the expected life of the 20 Richland
tanks equipped to accommodate self-boiling wastes is
probably no more than 20 years or could be as little as
10 to 15 years. Eleven of the 20 tanks have been in
service for 10 years or more. Further, recent studies
have cast doubt upon the wisdom of reusing such tanks
136
-------
after they have been emptied, regardless of their age.
In this regard, it appears that in the last half of 1969,
Richland may be confronted with a situation of having
only used tanks available as spare tanks for high-level
self-boiling waste storage; . . .
AEC informed us that the Lyons location probably could be
used for long-term storage of AEC's high-level radioactive
wastes; however, because of the estimated high cost (pre-
liminary estimates are in the range of $1.5 to 2 billion)
of processing, packaging, and shipping the wastes from
Richland and Savannah River to Lyons, efforts are under
way to determine whether suitable long-term-storage loca-
tions and methods can be developed at the two AEC sites.
AEC believes that the cost of exploring and developing a
long-term-storage method at these sites is justified,
because of the potential expense of shipping the large
quantities of waste at these sites to another location.
AEC advised us that it probably would not be economically
attractive for a commercial plant to make similar studies
for its own location. 14/
Scientists' concerns with nuclear waste have been cited in some length
because of the grave risks involved.
The Complacent View
Not all scientists have the same perception of nuclear waste disposal
problems. David Rose summarized problems related to nuclear waste management
this way:
I think that the nuclear waste problem can be solved rela-
tively cheaply; it has appeared as a problem mainly because
the former AEC tried to force the problem into a box that
was financially too small; anybody who would speak up
against the policy would best find employment elsewhere. I
have some personal experience with this particular matter.
That was unfortunate, because the problem is relatively
resolvable, at not too high a cost; there is a lot of option
space yet to be explored. 15/
Similar views are held by many other distinguished scientists.
Overview of the Study
These quotations, representing groups of opinions, indicate the extent of
polarization and conflicting opinions with regard to the societal risks re-
lated to nuclear waste management. They also indicate the applicability of
nuclear waste disposal as a case study for social decision-making under condi-
tions of low risk, high consequence outcomes.
137
-------
Chapter 12 presents a brief technical description of nuclear waste gener-
ated within the nuclear power cycle, and describes all major nuclear waste
disposal concepts and options. This provides a technical basis for discussing
societal risk related to nuclear waste management.
Chapter 13 discusses the risk issues associated with disposal of nuclear
wastes, and presents the more important unique physical features of nuclear
waste that affect risk issues. The basis for subsequent discussion of risk
analysis is derived from this information.
Chapter 14 gives the current legal framework for nuclear waste management
and current federal radiation protection criteria. These form the legal and
institutional basis for management of nuclear waste.
Chapter 15 contains a brief analysis of publicly voiced criteria for
nuclear waste disposal. These criteria are extremely important in assessing
strategies of nuclear waste disposal. They are uniquely important because
they tend to represent societal views of health related issues, both present
and future. The public participates in nuclear waste management decisions via
provisions embodied in the National Environmental Policy Act (NEPA). Thus,
understanding public concerns has become a critical prerequisite in analyzing
nuclear waste management.
Chapter 16 starts with a discussion of benefit-risk-cost analysis as it
applies to nuclear waste disposal. This chapter then uses the criteria de-
veloped in Part I of this study to evaluate the social decision process as it
functions with respect to management of nuclear wastes.
Notes to Chapter 11
1. Barry Commoner, The Closing Circle (New York: Bantam, 1972).
2. Edward Lawless, Technology and Social Shock, Final Report for National
Science Foundation (New Brunswick, N.J.: Rutgers University Press,
1977).
3. Gilbert F. White and J. Eugene Haas, Assessment of Research on Natural
Hazards (Cambridge, Massachusetts: MIT Press, 1975).
4. Pugwash News Letter, 11, July & October 1973, p. 23.
5. T. Hollocher, "Storage and Disposal of High Level Wastes," in: Union
of Concerned Scientists, The Nuclear Fuel Cycle; A Survey of the Public
Health, Environmental and National Security Effects of Nuclear Power.
Revised edition (Cambridge, Massachusetts: MIT Press, 1975), pp. 219-
220.
6. "Fission Energy and Other Sources of Energy," Science and Public Affairs;
The Bulletin of the Atomic Scientists, (January 1974), pp. 5-6.
138
-------
7. J. 0. Blomeke, J. P. Nichols, and W. C. McClain, "Managing Radioactive
Wastes," Physics Today. 26 (August 1973), p. 36.
8. A. S. Kubo and D. J. Rose, "Disposal of Nuclear Wastes," Science, 182
(December 21, 1973), p. 1209.
9. K. J. Schneider, "Solidification and Disposal of High-Level Radioactive
Wastes in the United States," Reactor Technology, 13 (Winter 1970-71),
p. 395.
10. Philip P. Micklin, "Environmental Hazards of Nuclear Wastes," Science
and Public Affairs; The Bulletin of the Atomic Scientists, 30 (April
1974), pp. 41-42.
11. Alvin M. Weinberg, "Social Institutions and Nuclear Energy," Science,
177 (July 7, 1972), p. 31.
12. Weinberg, "Social Institutions and Nuclear Energy. . ."pp. 33-34.
13. Allen V. Kneese, "The Faustian Bargain," Resources, No. 44 (Washington,
D. C.: Resources For the Future, Inc., September 1973), p. 1.
14. U.S. General Accounting Office, Observations Concerning the Management
of High-level Radioactive Waste Material (B-164052), May 29, 1968.
15. Proceedings of Quantitative Environmental Comparison of Coal and Nuclear
Electric Generators and their Associated Cycles Workshop, (Washington,
D. C.: MITRE Corporation, 1975), Vol. 1, MTR-7010.
139
-------
12. Disposal Concepts For Nuclear Wastes
This Chapter describes pertinent technical issues related to generation
and disposal of nuclear waste.
Generation of Wastes
Radioactive wastes generated by the nuclear fuel cycle can be generalized
into three categories: (1) high-level; (2) transuranium contaminated; and (3)
low-level. _!/
High-level wastes are generated from reprocessing highly irradiated nuc-
lear fuels. These wastes are so high in radioactive content of long-lived
isotopes that they require long-term storage in isolation with essentially
perpetual surveillance at storage sites. Before permanent storage can occur,
these wastes must be processed into inert, immobile, solid material which is
nonexplosive, noncombustible, and cannot turn to gaseous form and become air-
borne.
Transuranium-contaminated (TRU) wastes usually contain solid materials
contaminated with significant amounts of long-lived alpha emitters such as
plutonium.
Low-level wastes contain sufficiently low radioactivity that they do not
present significant environmental hazards. Other risks are implicit in radio-
active waste, however, regardless of the low level of contamination.
Although transuranium and low-level wastes present some initial risks, it
is high-level wastes that are of critical concern. The first high-level waste
was produced in the mid-1940's as a result of Manhattan Project activities.
Since then, such waste arising from defense production and nuclear power de-
velopment has been stored either as aqueous solutions or solids at various
installations. Waste has been stored as aqueous solutions at the only opera-
tional commercial fuel reprocessing plant (West Valley, New York).
Exponential growth of nuclear power in the United States will result in
increased quantities of high-level waste. The anticipated accumulation of
solidified high-level waste before the year 2000 is about 13,000 cubic meters.
Approximately 150,000 megacuries of radioactivity and 700 megawatts of heat
will be associated with this projected waste inventory in the year 2000. 2J
Figure 2 shows most of the potential routings of nuclear waste, starting with
its presence in the discharge reactor fuel and continuing to its final dispo-
sal.
141
-------
FIGURE 2
Generation of Nuclear Waste Within Nuclear Fuel Cycle:
Major Options for Waste Management
ro
3
-
Mining and
Milling
(
UF6
Production
1
Enrichment
<
Fuel
Fabrication
Reactor
Operation
Spent Fuel
Storage
i
Reprocessing
1
—
1
J
N
Trans
Wa
A '
orv
jranlc
stes *
Facilities
and
Lurge
Equipment
5 '
Transuranic
Wastes
6
High-Level
Waste
7
Spent
Fuel
8
Pu
*
— »•
—^-
3
Treatment
and
Packaging
1
Treatment
and
1
Solidification
and
Packaging
2 3
*-> _, . .
4
1 2 9
- 5™ ^ T«"W , . °*«*
i Interim
__ ., 2r~' Storage ' ~1 9
Inturim "
— Storaoe -*• TransP°ft _ 0l5P°Ml
Interim _____
1 Storaoe T
Packaging
*- Transport Disposal
__ Interim ______
._ . 1 Storaop t
Packaging
_,
* Lov-Lcvel Wastes
-------
The term "disposal" refers to placing the waste in permanent isolation,
sometimes supplemented by man-made barriers to provide adequate protection,
for a sufficiently long period for waste to decay to harmless radioactivity
levels. Short-lived waste fractions will decay and become nonradioactive in
relatively short times — that is, times sufficiently short to consider long-
term storage in man-made structures.
Partitioning (removal of actinides), fractionation (removal of certain
fission products), fixation or conversion to encapsulated solids, and retriev-
able storage are identified as processing steps of the waste management
method. Partitioning treatment of the waste is a key element in certain waste
management systems. Dividing the high-level waste into long- and short-lived
fractions allows the two waste fractions to be managed separately. The short-
lived fraction will diminish to low levels in about a hundred years.
To produce a short-lived waste fraction which would decay to negligible
radioactivity in about 1000 years requires separation of the actinide
elements and perhaps samarium, technetium, tin, iodine, and nickel (radioac-
tive nickel is due to irradiation and dissolution of some fuel cladding).
Large decontamination factors — the ratio of initial concentration in waste
to final concentration in waste — in a range 10" to 10& would be required for
some elements, particularly actinides, to render the remaining waste materials
non-radioactive after 1000 years of decay time.
In a broad sense, only three management options exist for the longer-
lived and highly toxic actinide fraction of the waste:
(1) elimination of waste constituents by beneficial transmutation, which
is nuclear conversion to other less dangerous isotopes;
(2) safe extraterrestrial transport off the earth;
(3) isolation from man's environment somewhere on earth for long periods
to allow natural radioactive decay.
However, as discussed in more detail in this Chapter, transmutation and extra-
terrestrial transport cannot be implemented at this time. Therefore, only
isolation is currently a viable alternative.
Characteristics of High-Level Waste
Table 6 shows typical constituents of the liquid high-level waste expec-
ted from the solvent extraction reprocessing of irradiated fuels from Light
Water Reactors (LWR), Liquid Metal Fast Breeder Reactors (LMFBR), and High
Temperature Gas-Cooled Reactors (HTGR). 3/ This waste is a nitric acid solu-
tion of chemical salts which typically contains a few volume percent of
solids. The waste constituents include: nonradioactive chemicals added dur-
ing reprocessing; almost all radioactive and nonradioactive fission products;
transuranium actinides formed in the reactor; and about 0.5 percent of the
fuel materials such as uranium, plutonium and thorium which are not recovered
143
-------
TABLE 6
Typical Materials in High-Level Liquid Waste
Reprocessing
Chemicals
Fuel Product
Losses—* JL'
Transuranic
Elements^/
Material-/
Hydrogen
Iron
Nickel
Chromium
Silicon
Lithium
Boron
Molybdenum
Aluminum
Copper
Borate
Nitrate
Phosphate
Sulfate
Fluoride
Sub Total
Uranium
Thorium
Plutonium
Neptunium
Americiura
Curium
Sub Total
Other Actinides—
Total Fission Products—
TOTAL
400
1,100
100
200
65,800
900
Grams/MT from Reactor
HTGR!/
3,800
1,500
400
300
200
200
1,000
40
6,400
40
435,000
1,100
1,900
Typei/
68,500
4,800
40
Sub Total 4,840
480
140
40
660
<0.001
28,800
103,000
452,000
250
4,200
1.000
5,450
1,400
30
1£
1,440
20
79,400
538,000
LMFBR-/
1,300
26,200
3,300
6,900
98,000
244,000
380,000
4,300
500
4,800
a/ Water content is not shown; all quantities are rounded.
b/ Most constituents are present in soluble, ionic form.
c/ U-235 enriched PWR, using 378 liters aq. waste/metric ton, 33,000 MKd/MT ex-
posure, (megawatt-days per metric ton of fuel)
d/ Combined waste from separate reprocessing of "fresh" fuel and fertile particles,
~ using 3,785 liters of aqueous waste per metric ton, 94,200 MV.VI/MT exposure.
e/ Mixed core and blanket, with boron as soluble poison, 10°,, of cladding dissolved,
~~ 1,249 liters per metric ton, 37,100 MWd/MT avg. expos.
f/ O.S°<. product loss to waste.
£/ At time of reprocessing.
h/ Volatile fission products (tritium, noble gases, iodine fj bromine) excluded.
144
-------
during reprocessing. Most chemical materials are added during fuel repro-
cessing. 4/
Present regulations 5J require that liquid high-level waste from fuel
reprocessing (1) be converted to a solid material within 5 years after sepa-
ration in the fuel reprocessing step, and (2) be encapsulated and shipped to
a Federal repository within 10 years of its production for long-term manage-
ment. Solidified high-level waste is assumed to be encased in steel canisters
averaging 12 inches in diameter and 10 feet long. Thus, solidification and
encapsulation must occur prior to initiation of most disposal schemes.
Four solidification processes have been developed in the United States to
the point of demonstration on an engineering scale:
- fluidized bed calcination
spray solidification
- pot calcination
- phosphate glass solidification
In all four processes, heat is applied to drive off volatile constituents
(primarily water and nitrates) which results in either a calcined solid or a
melt that will cool to a monolithic solid. The latter generally requires di-
lution of the waste with 20 to 40 percent nonradioactive materials such as
glass or ceramics that have low solubility in water and are fusible at tem-
peratures of less than about 1200°C.
Characteristics of typical final solid waste forms from the four pro-
cesses are shown in Table 7, and are described briefly below.
Fluidized Bed Calcination. Liquid waste is atomized into a heated fluid-
ized bed where it is deposited and calcined on granular bed particles. Re-
sulting granular "spheres" of waste calcine may be the final waste form or
they may be incorporated into crystalline or glassy solids in a melting stage.
Spray Solidification. Atomized droplets of waste fall through a heated
chamber where flash evaporation results in solid oxide particles. Glassmaking
solid frit or phosphoric acid can then be added to provide for melting and
glass formation. The molten glass or ceramic is cooled and solidified.
Pot Calcination. Liquid is continuously added and boiled away in a pro-
cessing vessel which also serves as the storage canister. When the canister
is full of solids, the addition of aqueous waste is stopped and the solid is
heated and held at about 900°C to complete denitration and dehydration. Feed
additives can be used to result in a glass rather than calcine cake.
Phosphate Glass Solidification. Liquid waste and phosphoric acid are
mixed and concentrated to a thick sludge in an evaporator. The sludge goes to
a melter where dehydration and denitration are completed and the material is
145
-------
TABLE 7
Characteristics of Solidified High-level Waste
Form
Description
Bulk Density (g/cm )
Kt?o Fission Product
Oxides (Max.)
Thermal Conductivity
W/(m2)(°C/m)
teachability in Cold
Water, g/cm -day
Pot Calcine
Scale
Calcine Cake,
Friable
1.2 to 1.4
90
0.3 to 0.4
1.0 to 10'3
Spray Phosphate
Ceramic
Monolithic
Ceramic
1.0 to 1.4
10"3 to 10"5
Phosphate
Glass
Monolithic
Glass
Hard, Brittle
2.7 to 3.0
25
0.8 to 1.2
-4 6b/
10 to 10 °~
Borosi licate
Glassi/
Monolithic
Glass
Hard, Brittle
3.0 to 3.5
SO
1.0 to 1.4
10"5 to 10"7
F 1 u i d i z ed
Bed Calcine
Granular
Calcine, Mean
particle diam.
100-500 micro-
meters
1 .0 to 1.7
SO
0.2 to 0.4
1.0 to 10"1
a/ Produced by either spray or fluidized bed calcining followed by melting, or by in-canister vitrifica-
~~ tion processing.
b/ Devitrified phosphate glass exhibits increased leachability (leach rates = 10 to 10 g/cm2-day).
-------
melted. Molten phosphate glass then drops into a storage canister where it
cools and solidifies.
Disposal Concepts
Table 8 summarizes three areas of basic disposal concepts—ocean, geo-
logical and ice sheet disposal. The following descriptions of these three
basic concepts, and others such as extraterrestrial and transmutation, have
been summarized from various Energy Research and Development Administration
(ERDA), Environmental Protection Agency (EPA), and Nuclear Regulatory
Commission (NRC) documents.
The concepts described here include the most frequently discussed and
proposed methods for disposing of radioactive waste materials. However, all
of these alternatives possess the same major elements of risk: engineering
failures; human failures; and unanticipated natural events.
Ocean Disposal Alternative^
Studies on ocean burial of radioactive wastes indicate that this concept
has considerable merit. _6/ Most importantly, many ocean regions are not only
isolated from human habitation, but are also biologically nonproliferating,
geologically stable, and without resources of societal need. Additionally,
areas have been identified which are isolated from natural disasters such as
earthquakes.
The deep sea floor, including ocean basins and abyssal plains, are con-
sidered geologically stable. In some areas waste could be isolated in bedrock
below the unconsolidated sedimentary cover. Large volumes of seawater would
help cool the waste and provide high dilution for any material v.'hich accident-
ally escapes. Additionally, the high ion-exchange capacity of sea floor sedi-
ments would provide additional isolation barriers.
\
Subduction zones/deep sea trenches are areas where, according to crustal
plate tectonics theory, one edge of certain crustal plates is moving under
other crustal plates, and then down into the earth's mantle. Radioactive
material could be buried in these trench areas and would be slowly subducted
into the earth's mantle with the crustal plate.
High sedimentation rate areas, where major rivers are building deltas in-
to the ocean, represent other sites where waste could be buried in the bedrock
below the accumulating deltaic sediments.
Solid radioactive waste management systems for seafloor disposal are dif-
ficult but manageable. Wastes could be encapsulated without prior partition-
ing of the different radioactive materials. Solidified and canned bulk waste
from the reprocessing plant would be packed in durable protective casks and
loaded aboard special ships at special embarkation points. A control system
at these checkpoints would maintain integrity of the waste canisters. Ships
would be equipped to provide cooling as well as monitoring and examination of
147
-------
TABLE 8
Summary of Nuclear Waste Disposal Concepts
Characteristics
Disposal Concept
Ocean Disposal
Direct dumping in ocean
Placement in ocean floor
in abyssal hills
regions
Placement in deep ocean
trenches for subduc-
tion
Ice Sheet Disposal
Geologic Disposal
Liquid Injection
Disposal in rained cavity
in salt
Disposal in deep holes
Extraterri'S trial Disposal****
Potential Practicality
Placement:
Risk
0* (C)
0 (0)
- (-)
- (-)
+ (+)
+ (+)
0 (0)
—
Stability
- (0)
+ (+)
- (0)
- (0)
- (0)
+ (+)
+ (+)
++
Isolation
- (o)
0 (+)
+*** (4.)
- (o)
- (0)
+ (+)
+ (+)
++
Invulner-
ability
- (0)
0 (+)
+*** (+)
0 (0)
- (0)
- (+)
+ (+)
-H-
Cost
+
0
-
_
+
0
0
--
Research
Needed
+
0
-
0
+
0
-
-
Existing
Tech.
+
0
-
0
+
0
-
-
oo
* _+ refers to positive attribute (high stability, low cost, etc.), 0_ to intermediate attribute, and -_ to a
negative attribute. Rating on basis of disposal on order of a million years.
** Values in parentheses are ratings on basis of disposal for 1000 years.
*** Ratings on basis that subduction has been achieved. Over much of life before subduction, rating would be
much lower.
**** Includes disposal of actinides only. Remaining fission products would have to be disposed of using another
method.
-------
the waste canisters until arrival at the disposal site. Holes in the basement
rock, about 800 meters in depth, would be predrilled on about 1.6 kilometer
centers to allow for heat dissipation. About 100 waste canisters would be
deposited into each predrilled hole. Disposal would be completed by filling
and sealing the top 200 meters of the holes with cement or grout and by re-
placing the sediment. Area surrounding the disposal holes would be monitored
for a specific period of time.
Since ocean burial would be an international issue, other countries in-
volved in producing nuclear energy might find a common dumping area ideal for
use among radioactive waste-producing nations.
Ice Disposal
Large permanent frozen masses of ice overlying continental land masses
offer potential advantages as a disposal medium. The Antarctic ice sheet,
which is International Territory, is a potential international repository.
Large areas of massively thick ice are available at locations remote from
human activities and with little chance of future development. Ice, provided
its average temperature remains far below the freezing point, self-heals frac-
tures. It has low permeability to water and a thermal conductivity comparable
to common rock types. Biological activity is low. The natural capacity of
the ice sheets to dissipate heat from radioactive waste canisters at low tem-
peratures is conducive to maintaining integrity of the waste materials.
Three potential disposal concepts have been developed for ice sheet areas
such as Antarctica or Greenland. These concepts, based upon managing the
solidified high-level waste without partitioning, are:
Meltdown or free flow. The waste canister is buried in an
individual shallow drilled hole in the ice and allowed to
melt down through the ice sheet to bedrock.
Anchored emplacement. The waste canister is placed in an
individual shallow drilled hole in the ice but remains con-
nected to surface anchors by cables or chains.
Surface storage/disposal. Waste canisters are placed in a
shielded cell storage facility with jack-up piers on the ice
sheet surface to provide heat removal to the ambient air and
to prevent covering by further ice accumulation.
For the meltdown and anchored emplacement concepts, the canister is
placed in a hole 50 to 100 meters deep. The canister descends by self-
melting. Cables or chains, 200 to 500 meters long and attached to surface
anchors, are tied to the canister for the anchored emplacement concept. These
cables control the waste canister's descent and maintain its position for the
required time period of about 100 years. Canisters are emplaced one per hole
on 1-kilometer centers. This spacing would maintain total separation between
canisters during descent. Disposal would be achieved by self-melting and re-
freezing of the ice above the canisters as they descend. The waste canister
149
-------
must be encapsulated to accomodate high pressures and prevent any possible
leakage of the waste.
In the surface storage facility disposal concept, storage would be
carried out by placing the canisters in a surface facility on the ice sheet.
After about 50 years, the facility is allowed to become covered by accumu-
lating snow and is eventually buried by accumulating ice and snow for final
disposal.
Transportation presents potential hazards. The waste management system
for all ice sheet concepts is characterized by transporting previously solidi-
fied and canned bulk waste in protective casks from the reprocessing plant to
special embarkation ports. These ports would contain a receiving facility
where the canisters would be inspected and placed in special protective casks
for loading onto the transport ship. Sea transport would be by specially de-
signed ship(s) with facilities for cooling, inspecting, and handling the
waste canisters. The hull would be designed for protection against damage due
to pack ice or related accidents. Ice-breaker escorts would be used for rou-
tine operation. The transport ship would operate outside normal sea lanes to
reduce the risk of collision at sea and would dock and unload the casks onto
the over-ice transport vehicles. The waste canisters and casks are then off-
loaded at a debarkation facility on land in an ice-free area near the edge of
the ice.
Crawler-type tractors pulling sleds are, at present, the most dependable
means of over-ice transport under the severe operating conditions of the ice
sheet areas. However, other types of transportation may be used such as air-
craft or surface effect vehicles (hovercraft), with fuel supplied by aircraft
fuel drops. Hovercraft would require further development and testing to prove
their capability for routine operation. Finally, the waste is emplaced at the
disposal site. Area around the disposal site would be monitored for the re-
quired time period.
Geologic Disposal
Problems of burying radioactive waste in remote geologic environments
have been intensively studied. Drilling, mechanical mining and dissolution,
hydraulic fracturing, and controlled use of explosives are some methods which
could be used. Table 9 shows ten conceptual methods for disposing of waste
radioactive material. All involve burial in environments far removed from
human and animal habitation for sufficient time to render the radioactive
waste harmless. These environments must be stable, with little or no possi-
bility of natural catastrophe such as earthquake or flood which might expose
the waste to the atmosphere. Naturally occurring elements such as thorium and
uranium have been isolated in this manner for millions of years. 7_/
Form of the material to be disposed is an important consideration when
studying geologic disposal methods. Solid waste appears to be the most feas-
ible. In several concepts, waste is buried in the geologic formation as a
150
-------
TABLE 9
Characteristics of Geologic Disposal Concepts
Concept
1
2
3
4
5
6
7
8
9
10
Type of Cavity
Mined
Mined
Mined-/
a/
Mined—
Mined
Exploded
Matrix of
Drilled Holes
Deep Hole
Deep Hole
Hydrofracture
Waste Form
at Time of
Emplacement
Solid
Solid
Solid
Solid
Liquid-/
Liquid-
Solid
Solid
Liquid^/
Liquid-/
Fluid
Cooling
No
Water
Air'/
Water-/
No
No
No
No
No
No
Waste- Rock
Reactions
No
Melt
No
No
Melt
Melt
No
No/Melt-/
Melt
No
a/ Includes underground man-made structures.
b/ All liquid emplacement concepts involve in-place conversion to a solid
form.
£/ Cooling is provided for an interim period of tens of years until the
heat generation rate has decreased to a point where melting will not
occur.
d_/ This deep hole concept is studied for both melting and non-melting
cases.
151
-------
liquid and then converted in place to a solid form for long-term disposal.
Under concepts 5, 6 and 9 in Table 9, liquid waste melts the surrounding rock
and, upon cooling, forms a solid waste rock matrix. Concept 10 indicates
liquid waste being incorporated within a self-curing cement.
In all of these concepts, transport presents serious and difficult haz-
ards when shipping aqueous high-level waste. Therefore, liquid emplacement
concepts require that fuel reprocessing plants be located at the disposal
site. Solid wastes are less hazardous to transport. If conversion to a solid
occurs at the fuel reprocessing plant, federal disposal sites can be at a
separate location from the reprocessing plant.
In addition to the general geologic disposal methods, numerous variations
have been proposed, the most important of which are reviewed below. _§_/
(1) Solid waste buried in mined cavity—no fluid cooling or melting^. Pre-
viously solidified waste is buried in the floor of rooms or tunnels excavated
in geologic formation. This approach is similar to the bedded salt disposal
concept studied over the past 15 years. Casks filled with radioactive wastes
are brought to the site by special carrier and lowered into a shielded hand-
ling cell. Individual canisters are removed from the cask and transported to
a shaft. Cables lower the canisters into the subsurface complex.
At the mine level, individual waste canisters are moved from the subsur-
face shielded cell into a shielded transfer vehicle. The subsurface transfer
vehicle receives the canister and transports the waste package to a previously
mined burial tunnel with predrilled holes in the floor. The transfer vehicle
is located over a hole, and the waste package is lowered into the hole. The
top 6 to 8 feet of the hole is filled with a sealant to provide isolation and
shielding. Heat from the waste is designed to be conducted through the geo-
logic formation without causing melting or deterioration of the geologic
media.
(2) Solid waste emplaced in mined cavities—interim liquid cooling and con-
version to rock-waste matrix. Waste canisters are removed from shipping casks
in a waste receiving facility, lowered through a drill hole into a lined cav-
ity, and deposited on the cavity floor in a random array. Waste within the
cavity is immersed in a boiling water bath. The steam which is generated is
condensed in a surface facility and returned to the hole for cooling the
waste. Operation of the surface cooling system is continued until the cavity
contents melt by allowing the cooling water to boil away. Cavity shutdown is
started by stopping return of the condensed steam. The cavity begins to dry
and the temperature of the contained materials rises. The cased hole is
sealed at an appropriate time, and the waste is allowed to melt. The waste is
concentrated in a small volume with little cooling by the surrounding rock and
will thus melt from its own heat within a few days. The heat from the waste
will melt some of the surrounding rock and form a larger molten mass which
dilutes the waste concentration. This process occurs over a 100-year period.
After a few hundred years, the molten rock-waste mixture will cool and solid-
ify into a rock-waste matrix as the heat provided by radioactivity diminishes.
152
-------
(3) Solid waste emplaced in man-made structures within geologic foundations —
interim air cooling. Inside the bottom of a thick-walled metal pod buried in
the floor of the tunnel, a canister of waste is sealed. Operating personnel
are shielded from radiation. Radioactive decay heat is conducted through the
metal pod wall and dissipated from the finned surface to the flowing cooling
air. The single pass cooling air flows by natural convection once the draft
has been initiated.
At the repository, a waste canister is unloaded inside a building and
transferred to the underground facilities. At the mine level, the waste can-
ister is placed in the containment pod. The pod is then placed in a hole or
trench within a tunnel. The hole is backfilled with crushed rock, and the
top of the pod is sealed. Other emplacement operations are similar to those
for concept (1) .
Solid waste buried in man-made structures in geologic formations — inter-
im water cooling. A man-made facility designed to withstand earthquakes and
shifting rock formations is tunneled underground. The waste canister is un-
loaded at the repository, transferred to the below-ground facility, trans-
ported to the disposal location through the waste canister handling area of
the tunnel, and lowered into the hole in a manner similar to the concepts
above. The waste is placed in the shielded lower part of a cylindrical water-
filled storage vessel in the mined cavity. Typically, the concrete shell
would be 30 feet in diameter; the tunnel, 50 feet in diameter. Steam gener-
ated by the waste is piped to a heat-exchange system at the earth's surface
where it is condensed and returned to the waste storage area. A side stream
of the coolant is treated to remove any radioactivity which may be present
from leaking or externally contaminated canisters. After a predetermined
time, the cooling water is removed from the system. The access areas are
sealed, and the waste heat is transferred to the surrounding rock without
melting, as in concept (1) above.
(5) Liquid waste buried in a mined cavity— in-place drying and conversion to
rock-waste matrix. Liquid waste is stored in a mined cavity below the fuel
reprocessing plant. The cavity contains a high-integrity liner or tank con-
nected through high- integrity piping to a specially designed condensing and
treatment facility for the radioactivity-containing vapor. This facility
would be located adjacent to the fuel reprocessing plant at the surface.
Liquid waste is added continuously to the cavity and cooled by recycling con-
densed vapors from the surface facility. When the reprocessing plant is to be
closed or the cavity is to be converted to its permanent disposal mode, the
recycle of the condensed coolant is stopped and the cavity contents are al-
lowed to boil to dryness and to melt. After assurance that no excessive pres-
sure will develop in the cavity, the access shafts and piping are sealed.
During a hundred-year period, the waste will melt the liner and some of the
surrounding rock. After several hundred years, the molten rock-waste mixture
will cool and progressively solidify into a rock-waste matrix.
(6) Liquid waste buried in exploded cavities — in-place drying and conversion
to rock-waste matrix. This concept, similar to concept (5), is based on the
formation of an unlined, rubble-filled cavity formed by nuclear explosion (or
153
-------
possibly by conventional explosive). Rock rubble filling the cavity furnishes
material for initial formation of the rock-waste matrix.
(7) Solid waste emplaced in a matrix of drilled holes—no mej-ting. All op-
erations in this concept are performed from the earth's surface. Canisters
full of solidified high-level waste are brought to the handling facility and
transferred to a specially designed transfer and emplacement (charging) ve-
hicle. The vehicle moves to a previously drilled hole where each waste canis-
ter is carefully lowered into the hole and placed on the preceding canister.
After the hole is filled with waste canisters to a predetermined level, the
hole is sealed, and waste is emplaced in the next hole of the array.
(8) Solid waste emplaced in a deep hole—in-place conversion to a rock-
waste matrix. Similar to the previous drilled matrix hole concept, this ap-
.proach has each hole drilled to an extreme depth, nominally 16 kilometers (10
miles) deep. The concept can be designed for non-melting or melting. For the
melting situation, the capability for handling vapors from volatilization of
small amounts of water in geologic formations may be required within the sur-
face facility. The waste in the lower portion of the hole will be molten
while waste is added from above. After each hole is filled to the predeter-
mined level with waste canisters, the hole is sealed. A few hundred years
after emplacement of the waste, the molten waste-rock mixture will cool and
solidify into a tall column of rock-waste matrix.
(9) Liquid waste emplaced in a deep hole—in-place drying and conversion to
a rock-waste matrix. This concept, also similar to concept (5), uses very
deep (16 kilometer) holes. The deep hole is drilled at the reprocessing plant
site. Liquid waste is continually added to the hole, and vapor from the hot
waste is removed in a vapor-condensing facility at the surface. A layer of
dried and calcined waste is located below the small pool of boiling aqueous
waste. Molten waste and melting rock form a layer below the layer of calcined
waste. When one hole is filled to the predetermined level (assumed in this
study to be the lower 7,500-meter portion of the hole), a second hole is used;
waste in the first hole is allowed to dry, and the hole is sealed. After a
few hundred years the molten waste-rock mixture will cool and form a waste-
rock matrix.
(10) Liquid waste emplaced by hydraulic fracturing—in-place conversion to a
solid. Liquid waste is mixed with cement grout to form a slurry which is
pumped into a well under pressure. The pressurized slurry spreads out to form
and fill thin horizontal fractures in the geological formation. These frac-
tures were previously initiated by forcing water at high pressure to frac-
ture the rocks along the bedding planes. The slurry cures in place within a
few days and forms a sheet of solid waste-cement mixture about 3 millimeters
thick and 350 meters in diameter around the well. Several layers of the
sheets of waste-cement can be emplaced one above the other, before sealing the
access hole and moving to another well location. Operations are all conducted
at the surface from a shielded cell facility. Dilution and heat transfer con-
ditions are designed to maintain the waste and rock at low temperatures (100°C
or less).
154
-------
Extraterrestrial Disposal and Transmutation
Both extraterrestrial disposal and transmutation offer considerable fu-
ture promise because of the low societal risk associated with their applica-
tion. In both concepts, the risks attendant to unanticipated natural events
are negligible. However, neither concept can be implemented at present, as
appropriate technology for these alternatives has not yet been developed.
Extraterrestrial disposal concepts. Assuming that a stable non-earth
intercept trajectory or orbit can be achieved, extraterrestrial disposal of-
fers complete removal of long-lived nuclear waste constituents from earth and
the potential for an international solution to waste management.
In considering extraterrestrial disposal concepts, disposal of only the
transuranic elements appears to be most practical because of the high shield-
ing weight of the disposal systems, the complex cooling systems required, and
the high space transport cost per unit of weight (at least $2000/kg of waste
material). Adverse features of this concept are: (1) only part of the waste
is considered; (2) possible launch safety problems; (3) retrievability and
monitoring are difficult; and (4) the concept would require international
agreements.
The overall waste management system would consist of interim aqueous
waste storage to allow for decay and simplification of partitioning. Parti-
tioning of the aqueous waste would result in a transuranic element fraction
contaminated by no more than 1% of the fission products, with remaining waste
to be disposed of by an alternative method. This actinide waste would be con-
verted at the reprocessing plant to a refractory oxide and then encapsulated
into high-integrity, multiple-barrier capsules. The capsules would then be
transported overland to a space launch site.
The waste would be launched into space to an initial low-earth orbit
(150 to 500 kilometers above the earth) with a reusable space shuttle. From
this orbit, a space tug or upper stage(s) would be launched to carry the waste
package to its final destination. In some cases, the launch system could in-
ject the waste to its final destination without subsequent course correction.
In other cases, the waste tug would require subsequent mid-course corrections
or propulsion. Monitoring would be required to control for off-standard
events and radioactivity in the upper atmosphere.
Transmutation elimination. Transmutation is the changing of one isotope
into another by any means. For waste management, transmutation results in a
product isotope having a lower toxicity and/or a shorter half-life than its
predecessor. Ideally, radioactive constituents in high-level waste could be
eliminated by using nuclear processes themselves to achieve the transmutation.
More practically, the transmutation process would accelerate decay of radio-
active waste by converting long-lived radioisotopes to other isotopes which
have shorter decay times. If this can be achieved, both the quantity of waste
containing long-lived radionuclides and the time required for safely storing
treated radioactive waste would be significantly reduced.
155
-------
The waste management system for waste actinides includes a period of
interim aqueous storage to allow for decay and to permit improved partitioning
into an actinide waste stream and a short-lived residue. At the reprocessing
plant, actinides are converted to oxides. These are then converted into spe-
cial recycle fuel forms. Finally, the actinide fuel is transported to a fis-
sion reactor for irradiation. The remaining short-lived fraction must be
disposed of by an alternative method.
Transmutation using fusion reactors may be feasible at some later time.
The concept is similar to the above, except that the aqueous waste stream
would be partitioned into two or more streams containing waste actinides, pos-
sibly selected fission products, and a residual waste stream to be eliminated
by other methods. The waste streams for transmutation would be converted into
solid fuel materials (probably oxides) at the reprocessing plant. Fuel ma-
terials would be fabricated into special target elements, which would be in-
serted into the blanket of a fusion reactor for irradiation as part of the
reactor fuel cycle. Continued recycle of material which undergoes transmuta-
tion would also be part of this operation. Waste streams not sent to trans-
mutation must be disposed of by other means. In addition, the "heel" of un-
transmuted actinide waste at the termination of a nuclear plant era must also
be rendered harmless.
A complete transmutation strategy for management of high-level waste as-
sumes a three-phase development approach:
— In phase 1, long-lived actinides are converted to short-lived fission
products. The normal nuclear fuel-cycle of the fission reactor industry re-
tains all actinides within this cycle and thereby converts a significant part
of the actinides into fission products by transmutation. Partitioning the
waste is the most significant modification to the fuel cycle.
— In phase 2, fission products are placed into temporary retrievable
storage, to be used in the next step in the process.
— Phase 3 transmutes in a fusion reactor the long-lived fission products
and actinides accumulated from the fission reactor cycle. This final phase
eliminates most long-lived radioactive waste constituents.
Notes to Chapter 12
1. T. Hollocher, "Storage and Disposal of High Level Wastes," in: Union of
Concerned Scientists. The Nuclear Fuel Cycle. . ., pp. 219-275.
2. H. P. Metzger, The Atomic Establishment (N.Y.: Simon and Schuster,
1972).
3. U.S. Energy Research & Development Administration, Alternatives for
Managing Wastes from Reactors and Post-Fission Operations in the LWR Fuel
156
-------
Cycle, Report coordinated by Battelle Pacific Northwest Laboratories
for the Division of Nuclear Fuel Cycle and Production, U.S. Energy Re-
search and Development Administration. (ERDA 76-43) 1976. Vol. 1, p. 9.
4. U.S. Nuclear Regulatory Commission, Environmental Survey of the Repro-
cessing and Waste Management Portions of the LWR Fuel Cycle, October
1976; NUREG 0116 (Supp. //I, to WASH-1248) , p. 4.24.
5. U.S. Energy Research & Development Administration, Alternatives for
Managing Wastes. . ., Vol. 3, p. 19.1.
6. U.S. Nuclear Regulatory Commission, Environmental Survey of the Repro-
cessing and Waste Management Portions. . ., p. 4.121.
7. U.S. Energy Research & Development Administration, Alternatives for
Managing Wastes. . ., Vol. 3, p. 20.1.
8. U.S. Energy Research & Development Administration, Alternatives for
Managing Wastes. . ., Vol. 1.
157
-------
13. Risk Issues Associated With Nuclear Waste Management
The alternative nuclear waste disposal concepts discussed in Chapter 12
were all characterized by essentially the same risk elements. This chapter
analyzes risk issues associated with nuclear waste management in more explicit
detail. It also discusses the unique characteristics of nuclear waste which
affect the risk features and social decision-making in the face of such risks.
Nature of Societal Risks Associated with Nuclear Waste
The principal societal risk resulting from nuclear waste is that of acci-
dental release of radiation which can affect all living organisms, specifi-
cally human health. The Biological Effects of Ionizing Radiation report
(BEIR Report) estimates that the cancer risk by radiation to an individual
organ such as the lung or bone is about 1 x 10~" cancers/year per rem dose to
the organ, and about 5 x 10 cancers/year per rem dose to the whole body. !_/
The BEIR results can be used to determine cancer risks. However, for continu-
ous exposure over a lifetime, which must be assumed if the ocean is contamin-
ated, the actual risk to an individual would depend upon type and timing of
exposure, as well as age of the individual at exposure. The BEIR Report sug-
gests that the cancer induction rate is maintained for a 30-year period begin-
ning 15 years after the irradiation (latent period).
In the case of exposure to bone from isotopes such as strontium and the
actinides, the 50-year dose commitment is received essentially continuously
over the individual's lifetime, meaning only about 1/50 of the dose is re-
ceived per year. However, this figure is compensated by the assumption that
people will continue to be exposed to the isotopes over their entire lifetimes.
Thus, in the 50th year of exposure, the yearly exposure rate will equal the
50-year dose commitment. When this observation is considered with data from
the cancer model proposed by the BEIR Report, and people of different ages are
considered (a uniform population distribution by age is assumed), the average
cancer risk per year for an average person from continuous uptake is 6 x 10~°
cancers/year per rem to the bone from a single year's uptake (using the ICRP
re values which are really 50-year dose commitment values); 20 x 10~« cancers/
year to other organs; and 100 x 10~6 cancers/year per rem to the whole body.
Using these values, the individual cancer risks for the three different time
scales can be summarized as follows:
- Short term:
- Intermediate term:
- Long-term:
6 x 10 cancers per year
6 x 10 cancers per year
_9
10 x 10 cancers per year
159
-------
Cancer risks noted above are only for those resulting from radiation dis-
charge in water. These figures represent the single highest risk from radia-
tion, however, and thus would also be the probable ultimate risk for radio-
active wastes released to other media. According to the BEIR Report, the
following four cases can be established to indicate the range of overall
health impact from nuclear waste radiation.
Case 1. Complete release of the total waste inventory, assuming
no fractionation or recycle of any waste materials, over
an average period of 1000 years.
Case 2. Complete release of the total waste inventory, assuming
fractionation and recycle of 99.5% of U, Pu, Np and 99%
of Am, Cm, I, Tc, over an average period of 1000 years.
Case 3. Complete release of the unfractionated waste inventory
over an average period of a million years.
Case 4. Complete release of waste fractionated as in Case 2 above
over an average period of one million years.
The above risks do not appear to be significantly greater than those
associated with many other human activities. Further, the above discussion
should clearly indicate that any societal risk from nuclear wastes results
only if radiation is released from nuclear waste dispositions. Factors that
may influence or cause such release are discussed below.
Factors That May Contribute To Radioactive Release From Nuclear Waste
As discussed in Chapter 12, all alternative nuclear waste disposal con-
cepts are intended to answer the physical integrity of the disposal site.
Three major factors that cause and/or contribute to radiation release from
nuclear waste deposition are:
- engineering and component failures
- human errors
- unanticipated natural events
These major factors are discussed in the specific evaluation of risk analysis
methodology detailed in Chapter 16. Secondary factors, of course, cannot be
ignored. Certainly there are events which could occur that would result in
release of radiation from geological disposal—geosphere to biosphere. A
schematic diagram of such events is presented in Figure 3. However, factors
that cause and/or contribute to release of radioactivity from nuclear waste
depositories are relatively few in number and conceptually not difficult to
analyze and assess. The complexity in nuclear waste management and in assess-
ing the societal risks results from the unique characteristics of nuclear
waste.
160
-------
FIGURE 3
Geological Disposal From Geosphere Migration Point Of View
GEOLOGIC DISPOSAL
TYPES OF
FORMATIONS
NON-SALT
FAULTED MONOLITH
NON-SALT
PARTICULATE MEDIA
ETC
PATHWAYS FROM GEOSJ
I
SALT
PARTICULATE MEDIA
ETC
'HERE TO BIOSPHERE
UPHEAVAL
LEACH
EXCAVATION
EARTHQUAKE
\
f
MODEL
N
GLj
/
MODEL M
V
f
MODEL
BIOSPHERE
PATHWAYS
*Source: Bradishaw, R. L. & W. C. McClain, "Project Salt Vault: A Demonstra-
tion of the Disposal of High-Activity, Solidified Wastes in Under-
ground Salt Mines," ORNL-4555, 1971.
161
-------
n_a_r a_c_t_e_r i_s_t ics _o f N uc 1 e_ar_Wast_e_
Two characteristics c.onl'rlbute to analytic and data difficulties of
nuclear waste management more than any other. First, nuclear waste remains
potentially hazardous for a very long time span, thereby creating severe
intergenerational effects. Second, certain waste management decisions are
irreversible.
The time during which the waste remains active is indeed enormous, span-
ning hundreds of thousands of years for long-lived wastes. The magnitude of
this span can be better comprehended by comparing the required control time
for nuclear waste management with other characteristics of natural ages, as
shown in Figure 4. Further illustration of this long time span can be seen
from the curves that indicate the hazard index of spent fuel as a function of
age, shown in Figure 5. Because management decisions regarding nuclear waste
are without precedent in managing public policy, this lengthy time span must
be recognized and considered, creating formidable conceptual, methodo-
logical and data problems that render most benefit-cost or benefit-risk analy-
sis unreliable in relation to nuclear waste management.
An additional consideration is the problem of setting appropriate inter-
est rates in benefit-cost analysis (discussed in more detail later in this
report). The root of this complexity lies in one component of the real
interest rate: risk premium. As stated, the risks of nuclear waste disposal
can be large for all of present and future society. It is extremely difficult
to derive the shape of utility functions toward risk for future generations
over the span of this long time period. Therefore, long-term benefit-cost
analysis cannot be used since the results of such analysis are highly sensi-
tive to the interest rate.
Two additional elements that contribute to the difficulties of under-
taking meaningful benefit-cost or related analysis of nuclear waste disposal
are the relatively complicated legal framework of nuclear waste management,
and the ambitious, albeit incomplete, Federal Government radiation protection
criteria. Analysis of both of these elements and their relationship to analy-
sis of societal risk management appears in the following Chapter.
Note to Chapter 13
1. National Academy of Sciences-National Research Council, Division of Med-
ical Sciences, Effects on Populations of Exposure to Low Levels of Ion-
izing Radiation.Report of Advisory Committee on Biological Effects of
Ionizing Radiation . (Washington, B.C.: National Academy of Sciences,
Tlovember 1972). This report is frequently referred to as the "BETR
Report."
162
-------
FIGURE 4
Comparison of Characteristic Natural Ages with Control Time of High-Level Waste
High-Level
Wastes -
Control Time
to Hazard of
Natural Ores
Continents
Oceans
Antarctic
Ice
Man
Iodine § Presently
Actinides Placed Waste
Removed Actinides Removed Composition Spent Reactor Fuel
til i i • i i i
Age of Age
Age of Salt Basement of
Deposits Rocks Earth
Residence Time Period of
of Deep Ocean Stability of Oldest Ocean
Water Continental Shelf Floor Rocks
Existence
Residence Time of
of Ice Ice Gap
Lifespan Age of Existence
of Man Pyramids of Man
^^^^^^^^^"™* j -j i t
i * .. — A • i < r 1 1 1 —
10'
10
10"
10
10"
10C
10
10
10
10
10
11
*Source: General Accounting Office, "Isolating High-Level Radioactive Waste from the Environment:
Achievements, Problems, and Uncertainties," RED-75-309, December 1974.
-------
FIGURE 5
Hazard of Spent Fuel as a Function of Age
JYpica1 Uranium Ore
10
10
10J 10" 10' 10*
AGE of WASTE - YEARS
*Source: Gera, F. and D. G. Jacobs, Consideration in the Long-Term Management
of High Level Radioactive Wastes, U.S.A.E.G. Report, ORNL 4762, Oak
Ridge National Laboratory, Tennessee, 1972.
164
-------
14. Legal Framework and Federal Criteria
for Nuclear Wastes
This chapter gives a brief account of the legal framework for nuclear
waste disposal plus a detailed discussion of current Federal Government radi-
ation protection criteria. These two subjects, in the context of waste dispo-
sal concepts, have differing impacts on analysis of risk. Impact of the legal
framework is quite explicit and clear. Conversely, because radiation protec-
tion criteria are still in a state of evolution and refinement, their impact
is less clear and may undergo considerable change in the future. Neverthe-
less, these important criteria cannot be neglected.
Legal Framework for Nuclear Waste Management
Radioactive waste is one group of a large number of waste materials con-
sidered hazardous. Several federal and state statutes pertain indirectly to
management of radioactive waste through controls established for hazardous
wastes management. Other federal statutes, such as the Atomic Energy Act of
1946 and Amendments in 1954 and the Energy Reorganization Act of 1974, have
more direct impacts on management of radioactive wastes. The Supreme Court
has ruled that states may not regulate emissions from nuclear plants. ~\J As a
result, uncertainty exists regarding the extent to which states can directly
affect management of radioactive wastes.
For many years the most important government agency affecting development
of nuclear energy and, consequently, management of radioactive wastes, was the
Atomic Energy Commission (AEC). This Commission was established by the Atomic
Energy Act of 1946, and abolished in 1974 by the Energy Reorganization Act.
The AEC functioned as both promoter and regulator of nuclear energy. In this
dual capacity, the Commission was thought by some to represent a conflict of
interest, 2j which later proved to be one cause for its abolishment. The
Energy Reorganization Act established two agencies: (1) the Nuclear Regula-
tory Commission (NRC), which was given the regulatory function of the AEC; and
(2) the Energy Research and Development Administration (ERDA), which assumed
the promotion function of the AEC. Subsequently, the Department of Energy
(DOE), created by the Carter Administration in 1977, absorbed ERDA.
The Atomic Energy Commission. The Atomic Energy Act of 1946 established
the Commission to control development of atomic energy. 3/ At that time
atomic energy held great potential as a source for electrical energy and for
national defense. Accordingly, the Commission was given enormous powers, un-
precedented in American history. The scope of its powers was suggested by the
165
-------
legislation's specifications that the Commission alone could own certain
nuclear materials, kj
In 1954, the Atomic Energy Act of 1946 was amended to allow private in-
dustry and the public to participate in development of nuclear resources for
production of electrical energy. The AEC's role was relaxed slightly from
total control to that of promoting and regulating development of nuclear en-
ergy. It became possible for private industry to own reactors and other fa-
cilities, all of which were subject to licensing and other regulations imposed
by the AEC.
AEC's purpose was to assure that nuclear energy made a maximum contri-
bution to national defense and contributed to the general welfare. To achieve
these goals, the Commission developed programs:
a. To conduct research projects and assist in related
projects initiated by government and private enter-
prise.
b. To disseminate unclassified information and to de-
classify certain restricted data for dissemination.
c. To encourage widespread participation in development
and utilization of nuclear energy for peaceful pur-
poses to the maximum extent consistent with common
defense and security, and with considerations for
public health and safety.
The AEC was responsible for granting, suspending, revoking, or taking
other appropriate action regarding the licensing of facilities, issuing of
construction permits, or transferring a license from one party to another. In
1961, the AEC established an office of the Director of Regulations, and di-
vided its staff into two units: regulations and operations. Heads of both
units, the Director of Regulations, and the General Manager, reported directly
to the Commission.
Applications for licenses or amendments to existing licenses had to be
filed with the Director of Regulations. Separate licenses were required to
construct and operate a nuclear facility. Parties intending to file applica-
tion for either kind of license had the option of conferring informally with
the regulatory staff prior to filing such application.
The Atomic Energy Act required that each application for a license con-
tain all information that the Commission deemed necessary to reach a decision.
These requirements included information on the financial and technical quali-
fications of the applicant, as well as character and citizenship. In addi-
tion, the application had to contain technical information, such as the kind,
quantity and source of the nuclear material required; intended location of the
facility; specific characteristics of the facility; and a preliminary safety
report (PSAR).
166
-------
Applicants for a license to operate a power plant or radioactive waste
management facility had to provide the chief executive of the municipality or
the county in which the facility was to be located with a copy of the appli-
cation filed with the AEC. The Director of Regulations also sent a copy of
the application to the Governor of the State in which the facility was to be
located and arranged to have a receipt of the application published in the
Federal Register.
The regulatory staff, with assistance from the operating staff, reviewed
the application to determine whether the proposed facility complied with
national security, public health and safety goals of the AEC. The staff
had to follow the Administrative Procedure Act, excluding the requirements for
classified information. 5J The Justice Department had to be consulted for
advice on antitrust matters. _6/ The Environmental Protection Agency had to be
solicited for advice on public health. ]_/ In addition, the Commission had to
follow the procedures and implement policies of the National Environmental
Policy Act (NEPA). 8/
Usually, the first problems encountered by the Regulatory staff involved
inadequate information. Moreover, the applicant's evaluations of certain as-
pects of the facility often differed from the staff's evaluations. 9f The
staff would attempt to resolve these differences through conferences and writ-
ten communication with the applicant.
The Advisory Committee on Reactor Safety (ACRS) would also review the
application regarding public health and safety, and would report its findings
in the form of a letter addressed to the Chairman of the AEC. The Regulatory
staff would use all of the information in preparing a Final Detailed Statement
on Environmental Considerations.
Prior to issuing a construction permit, a mandatory public hearing was
conducted by the Atomic Safety and Licensing Board. Another public hearing
was necessary prior to issuing an operating license if the licensing was con-
tested. Final decisions of the Commission to issue either a construction per-
mit or an operating license were reviewable in a Court of Appeals. 10/
The public hearing was the AEC's official mechanism for public input.
The scope of each hearing was established by the AEC. Intervenors often re-
garded the scope as too restricted and raised issues, such as the effects of
waste disposal and alternatives of purchasing electricity elsewhere or using
less electricity, but the relevance of such issues was not finally
resolved, ll/
Another problem perceived by intervenors concerned the AEC's rule for
implementing policies of the National Environmental Policy Act. In the land-
mark case of Calvert Cliffs' Coordinating Committee vs. AEC _12_/, the Court of
Appeals for the District of Columbia found that the Commission's interpreta-
tion of the NEPA improperly limited consideration of environmental values in
the Commission's decision-making process. 13/ Prior to this case, water qual-
ity questions could not be considered at the hearing, and other environmental
effects were considered only if contested. 14/
167
-------
The AEC prohibited any construction at nuclear power plant sites until a
construction permit was issued. This prohibition banned all activities which
would adversely affect the natural environment at the site, including cleaning
of land, excavation, or construction of non-nuclear support facilities. 15/
Certain other activities, such as preconstruction monitoring to establish
background information related to suitability of the site or to protection of
environmental values, were permitted. This included geologic, seismic, hydro-
logic, and meteorologic investigations, as well as building such roads and
physical structures as were judged reasonably necessary for determining site
suitability. 16/ Once construction of the facility began, the AEC also had
the authority to order changes to the facility while under construction, if
such changes were economically justified. 17/
The Atomic Safety and Licensing Board stipulated the conditions under
which the facility would operate, in accordance with technical analysis
of the safety features. The Board usually set upper and lower operating
limits on the facility and a maximum number of years over which the plant
would be operated, not exceeding 50 years. 18/
Other Federal legislation. As noted previously, regulatory and promo-
tional functions of the AEC have been taken over by the Nuclear Regulatory
Commission and the Department of Energy, respectively. Other federal agencies
with a role in radioactive waste management include the Departments of Trans-
portation (DOT), Health, Education, and Welfare (HEW) and Environmental Pro-
tection Agency (EPA). Applicable statutes are (1) the Transportation of Ex-
plosives Act which prohibits transport of radioactive materials unless the
public interest requires expedited movement or such transport poses no danger
to life or property 19/; (2) the Hazardous Materials Transportation Act of
1970, and (3) the Hazardous Cargo Act which regulates water transport of
explosives or dangerous substances. 20/
The Federal Clean Air Act authorized the Administrator of EPA to set
standards on hazardous air pollutants at any level "which in his judgment pro-
vides an ample margin of safety to protect the public health." 21/ Finally,
the Federal Water Pollution Control Act contains several provisions which
directly affect management of radioactive wastes.
Legislation with the most critical impact on any analysis is, of course,
the National Environmental Policy Act of 1965 (NEPA). The importance of NEPA
for nuclear waste management results from requirements for execution of a
detailed comprehensive benefit-cost analysis. The Act also permits interven-
tion during the planning stages of nuclear waste deposition. Recent events
clearly indicate the importance of such intervenors in public policy proces-
ses, and their presence adds measurably to the risk analysis requirements.
These requirements are discussed further in Chapter 16.
As described above, several federal agencies have jurisdiction over
nuclear waste disposal. This often makes interagency agreements necessary.
Due to lack of a clear-cut division of responsibilities and authority, dis-
pute occasionally arises regarding jurisdiction as well as responsibility.
Inherent in this situation is a sometimes unnecessary duplication of effort.
168
-------
Current Federal Government Radiation Protection Criteria: Their Applicability
To Nuclear Waste Disposal
This section analyzes the most recent federal environmental protection
criteria, as issued by the Environmental Protection Agency, which are most
applicable to the guidance of radiation policy for operation of nuclear waste
disposal activities. 22/
The Environmental Protection Agency is responsible for establishing "gen-
erally applicable environmental standards for the protection of the general
environment from radioactive material." 23/ In this context the term "general
environment" means "outside the boundaries of locations under the control of
persons possessing or using radioactive materials." Promulgated standards
specifically exclude waste management operations, and thus avoid certain tech-
nologically defined aspects of these operations. However, they serve as
generic models for waste disposal protection criteria to be issued at a later
date. 24/ It is important the the Environmental Protection Agency have these
"generic models" on hand for assistance in formulating specific waste manage-
ment criteria. Moreover, present criteria will probably require only "minor
perturbations" to be applicable to the waste management portion of the fuel
cycle. 25/
Ultimate objectives. A statement of objectives is a useful administra-
tive guide for development of more specific criteria. Ultimately, these cri-
teria will be refined to a point where decisions can be based upon them, and
operations can proceed with a high degree of expected success. Objectives
broadly defined are as follows:
(1) To protect the public against radiation doses resulting
from fuel cycle operations.
(2) To limit the environmental burden of long-lived radioactive
materials that accumulate as a result of producing electri-
cal energy, so as to limit their long-term impact on both
current and future generations. 26/
The first objective is to develop a standard which sets limits on annual
doses of radiation. The second objective requires a standard to limit the
burden of radioactive materials, the stock of the cause of risk. Final stan-
dards parallel this division of concern into the flow (annual dose) and stock
of risk. Precise quantitative limits on annual dose and quantity of certain
long-lived radioactive elements (in terms of their release per gigawatt-year
of electrical power) are specified. Eight criteria used to implement, refine
and modify the above objectives are discussed on pages 170-174, following the
next subsection on secondary objectives.
Secondary objectives. Two further considerations have guided development
of standards. First, a report by the National Academy of Sciences contained
the judgment that "the current Guide is too high." 27/ This assessment was
accepted by the Environmental Protection Agency, and resulted in more conser-
vative standards. Second, language embodied in the Radiation Protection
Guidance as well as many Congressional statutes requires that exposure be kept
169
-------
as "low as practicable." A judgment was made that this must be interpreted to
mean that consideration could be given not only to radiation exposure, but
also to costs of controls. Thus, from onset of the standards development pro-
cess, dual and possibly opposing objectives existed: to lower the present
radiation standard, and to consider the costs as well as the benefits or re-
duced risks of levels of exposure.
Eight criteria. Note that the decision process can be characterized in
part as a cost-risk-benefit analysis. To some, cost-benefit calculations im-
ply maximization of a single, well-defined objective function. Actually, the
decision-making process is considerably more complicated. Various criteria
have been developed and used to make decisions involving a number of trade-
offs among different objectives. Thus the decision process can be character-
ized as one of finding a set of standards that satisfy multiple criteria.
The most important criteria are discussed here.
Criterion 1: Human Health
Radiation can harm all living species: mankind, animal life, and
plants. Although all living organisms may be affected by radiation, standards
were calibrated in relation to their effect on man alone, since "there is no
present evidence that there is any biological species whose sensitivity is
sufficiently high to warrant a greater level of protection than that adequate
for man." 2_8/ Thus the analysis estimates potential health effects to humans
as the measure of risk, rather than more general measures of quantities of
radioactive materials.
Criterion 2; Consideration of Planned Releases Only
Since the Environmental Protection Agency is responsible for establishing
performance standards, and the Nuclear Regulatory Commission is responsible
for establishing operating standards to ensure accomplishment of performance
standards, hazards of uncontrollable events are excluded from consideration.
Control of normal operations to reduce planned effluents is only minimally
"coupled" with the potential for accidental release of radioactive material.
A separate process addresses the issue of accidental releases. A variance on
the standards is proposed in case of "an emergency need for uninterrupted
delivery of power," or "in the presence of a temporary and unusual operating
situation." 29/
Criterion 3; Reduction pf_TptalPotential Health Impact on Large Popula-
tions
The most important criterion is the aggregate dosage affecting various
populations on a local, regional, national, or worldwide basis. 30/ Future
and present populations are to be considered. This aggregate goal, which can
be compared to maximization of national income in classical cost-benefit
analysis, does not consider distributional effects. Within a given level of
performance of this criterion, it does not matter whether the committed doses
are shared equally or impact certain members of the population much more
heavily than others. A central tenet of classical cost-benefit analysis is
that such distributional considerations should be secondary. Undesirable
170
-------
distributional effects (at least those concerning money) can always be
handled through income redistribution, provided the aggregate efficiency cri-
terion is satisfied. However, when the analysis concerns health effects, it
is much more difficult to claim that an avenue of compensation exists. Health
is much less fungible than money. In general, a healthy person cannot give
away part of his health to another.
The criterion as stated is unclear regarding relative weights to be as-
signed various populations. Whether people of the United States bear more
heavily than other people in the world is not stated. More importantly, no
balance is established between present and future generations.
Criterion 4: Distributive Effects Must Be Considered
Even though the aggregate dosage criterion is satisfied, an undesirable
distribution may exist. 31/ Since the standards are environmental rather than
occupational, this criterion addresses neighbors of particular site operations
who may receive higher than average doses. Radiation emissions are distribut-
ed more heavily around nuclear operation sites, and it is therefore reasonable
to expect that radiation doses will be distributed more heavily around sites.
Since the incidence of health effects is assumed to be proportional to dosage,
the total number of health effects should be invariant to the distribution of
radiation doses. 32/ Given a level of aggregated health effects which satis-
fies Criterion 3, no health gains can be made by redistributing the burden of
the dosage. An unfair situation is thought to exist, however, when some per-
sons are exposed to much higher risks than others, particularly "when the
individual at risk is not the direct recipient of the benefits of the activity
producing them." 33/ This notion accords with two commonly expressed prin-
ciples of equity.
First, society should be willing to pay far more to avoid involuntary
risks. As individuals, neighbors of nuclear plants for the most part do not
choose to live near a source of radioactive materials. Society should be
willing to pay more to reduce risks to these persons than it may decide to
pay for avoidance of risks to the United States population as a whole, who
presumably have voluntarily accepted the risks of nuclear power.
A second implied principle of equity is less common. Many economists
believe that persons working in more dangerous occupations demand and receive
risk premia — higher salaries in compensation for the higher level of risk
on the job. Since these standards are environmental, belief in the equity of
risk premia as fair compensation for more hazardous occupations is not a basis
for accepting a maldistribution of risks from nuclear materials. The neigh-
bors of atomic plants are not compensated for living near such plants. The
benefits of nuclear power accrue to society in general, not to purchasers of
the electricity produced by nuclear power, and certainly not to the neighbors
of nuclear and fuel cycle operations.
If effects to individuals must be considered, the objective of minimizing
the maximum individual dose alone might be considered. In this connection,
economic factors must also be considered. Otherwise, the standards might
171
-------
"achieve negligible improvements in public health protection for unreasonably
large investments in control technology." 34/
Criterion 3 indicates that the sum of overall health effects is to be
minimized, giving due care to the distributive burden of radiation by Criter-
ion 4. Answers are needed as to how far minimization of individual dose
should proceed. These answers, in part, are provided by the next criterion.
Criterion 5: Reduction of Risks Must Be Cost-Effective
As previously mentioned, a secondary objective of the standard setting
process was to consider not only minimization of radiation exposure but also
cost of controls. The development of implicit criteria for the requisite
cost-effectiveness trade-off is the primary subject of the following dis-
cussion.
Explicit guidelines to cost-effectiveness are not usually present in the
final standards. Such guidelines are not necessary if the decision-making
process is considered as something other than a cost-risk-benefit analysis.
However, the polar dangers of ignoring cost and risk trade-offs are explicitly
recognized. Excessively high radiation standards would expose the public to
"unnecessary" health risks — "unnecessary in the sense that exposures per-
mitted by the standards can be avoided at a small or reasonable cost to the
industry." 35/ On the other hand, standards would be too low if their control
costs were "unreasonable in the sense that control costs imposed by the stan-
dards provide little or no health benefit to the public." 36/
Further, two possible options to avoid the difficult task of making cost-
effectiveness choices are also discarded. First, it is recognized that no
scientifically determined "acceptable" or "justifiable11 level of health ef-
fects currently exists. Were such a criterion available, the decision process
would be far more simple: only the technological task of determining "which
controls should be used in order to meet that criterion at lowest cost" would
need to be resolved. 37/ But, for the reason stated, cost minimization can-
not be followed. Risk minimization is a second possible option. If only a
specific amount of money could be spent for health protection, there would
remain a relatively simple, value-free process of minimizing the number of
health effects. 3_8/ However, neither an absolute health nor cost criterion
exists, so the value-ridden trade-off between costs and permitted health ef-
fects must be made.
The Environmental Protection Agency approaches this problem by discussing
several subordinate considerations that cannot be ignored. 39/ Since these
considerations are classical constraints on choices in cost-benefit analysis,
they are subsumed under the generic criterion that the reduction in risks must
be cost-effective.
Criterion 6; Streams of Economic Expenditures Over Time Should Be Dis-
counted to Present Values
In the analysis of various control options that might be required as a
result of the proposed standards, capital and operating costs over the
172
-------
expected life of each system were determined as present values. This is com-
mon practice in cost-benefit analysis. Since purely economic resources are
fungible over time, and since future costs to society are less burdensome than
the same costs today, the discounting procedure seems unexceptionable. How-
ever, the question of the proper discount rate is left unmentioned. This cri-
terion is thus incomplete.
Criterion 7; Streams of Health Effects Over Time Are Not Discounted,
but Instead Are Summed with a Cut-Off at 100 Years
Although the standards are not intended to cover the major portion of the
nuclear fuel cycle with long lasting effects (the disposal of wastes) , they
are applicable to a number of elements which will expose the population to
risks over periods much longer than 100 years. The analysis includes consid-
eration of the dangers from long-lived radionuclides:
The total significance of (the) environmental burdens of
carbon-14, iodine-129, and the long-lived transuranics ,
which have half-lives of 5700 years, 17 million years
and from 18 years to 2 million years, respectively, can-
not be quantitatively assessed, but must be assumed to
be greater than that anticipated during the first 100
years alone.
It is also recognized that the costs society is willing to pay to avert
health effects depend in part on the timing of these effects:
"The amount depends upon. . .how far into the future
it (loss of human life) is anticipated to occur." 4l/
Many economists would find serious flaws in the above criterion and its
justification. There is an inherent contradiction in discounting operating
costs but summing health effects. 42, 43/ Economic theory would concur that
neither individual lives nor health are fungible or marketable. However,
transactions exist which expend monetary resources and obtain increases in
health or longevity. Hence money — in a marginal sense — is convertible
into life and health. Society today spends money for hospitals, physician
care, and medical research. This money can be saved, invested, and a greater
dollar amount of similar health care purchased at a later time. In this
sense, through the market back-door, economic equivalents to life and health
exist and can be discounted.
What is inherently contradictory is to place the same (non-monetary)
value on a life saved in 100 years as a life saved today, while completely
neglecting the value of a life saved in 101 years. In fact, the results of
the contradiction, though theoretically confused, accord with the views of
economists who would have the analysis discount monetary equivalents of all
lives saved in perpetuity.
If an upward bias is placed on the valuation of the stream of lives by
not discounting, a downward bias is imposed by not counting the value of lives
after 100 years. Given the state of uncertainty concerning expected health
173
-------
effects on lives after 100 years, the net effect of the two biases cannot be
precisely determined. Given a constant valuation for life or health effects,
however, the procedure of using undiscounted sums for 100 years is equivalent
to summing discounted lives or health effects in perpetuity at some implicit
discount rate. The implied discount rate can be shown to be quite low, prob-
ably well below that used to discount the monetary costs in the analysis. 44 /
The effect of Criterion 7, therefore, is to weigh health effects more heavily
than if economic costs and health effects were handled symmetrically, both
using the same discount rate.
Criterion 8; Choice of a Cost-Effective Level of Control is Jointly
Dependent on Technological Circumstances and Societal
Valuations of Health and Life
Having avoided the issue of valuing life or health in computing the bene-
fits of radiation reduction, the issue must rise anew when a choice on a
health effects-cost trade-off curve must be made. The relationship between
radiation control costs and reduction of health effects could be charted. For
example, the health effects per gigawatt-year of electricity could be plotted
against the cumulative present values of additional control measures, with
this latter scale translated into incremental cost to the consumer in terms of
mills per kilowatt hour. 45/ Health effects per additional million dollars of
control equipment could then be plotted against the cumulative present worth
and incremental costs to the consumer. 46/
At a certain point on a. chart of this nature, "a breakpoint would occur
between efficient and inefficient control options." 47/ The marginal cost for
reducing potential health effects might be roughly $500,000, say, and a judg-
ment then made that beyond this point "an insignificant further reduction in
health effects is obtainable even for large additional control expendi-
tures." 48/ The terms "insignificant" and "large" are, of course, defined
only by values of the Environmental Protection Agency decision makers. Choice
of a cost-effective level of control would appear to have been selected by
examination of when the trade-off curve pitched steeply downward, and then
further justified by a comparison with other estimates of the proper amount of
resources to be spent for preservation of potential statistically defined
lives.
A cost-effectiveness approach such as that just described does not find
general agreement in other government studies and decisions. Moreover, no such
cost-effectiveness level exists. Finally, cost-effectiveness cannot be a cri-
terion for determining the degree or level of control.
Summary
The legal-regulatory framework within which decisions concerning disposal
of nuclear wastes have been made was reviewed briefly. One of the more impor-
tant developments has been the National Environmental Policy Act of 1965,
which greatly enhanced the role of cost-risk-benefit analysis. This act, in
conjunction with certain court decisions, has forced regulatory authorities to
base nuclear decisions in a considerably broader context than before.
174
-------
Eight criteria developed by EPA for radiation hazards in the environment
were also reviewed in some detail. These criteria have been used to develop
specific studies for environmental exposure to radiation. Although not speci-
fically developed for nuclear wastes, at the time this study was undertaken
they were the only environmental criteria in existence. Hence, they may well
form the basis for subsequent development of criteria dealing explicitly with
nuclear wastes. These criteria contain certain internal inconsistencies, as
was pointed out, and in certain ways they appear somewhat general. However,
as will be seen in the next chapter, they are far more detailed and specific
than any criteria advanced to date by public interest groups.
Notes to Chapter 14
!• Northern States Power Co., v. Minnesota, Federal Reporter, 447F2d 1143
(1971), pp. 1143-1158.
2. Northern States Power Co., v. Minnesota,. . .
3. U.S. Atomic Energy Commission, Rules and Regulations, Title 10-Atoroic
Energy, Part 2, Rules of Practice, October 27, 1966, pp. 21-36.
4. W. C. Reynolds, The California Nuclear Initiative; Analysis and Dis-
cussion of the Issues (Stanford, CA: Stanford University Institute for
Energy Studies, April 1976), p. 5.
5. Association of the Bar of the City of New York, Electricity and the
Environment: the Reform of Legal Institutions, Special Committee on
Electric Power and the Environment, Sheldon Oliensis, chairman. (St.
Paul, Minn.: West Publishing Company, 1972), p. 69.
6. Association of the Bar of the City of New York, Electricity. . .
7. Association of the Bar of the City of New York, Electricity. . .
8. Association of the Bar of the City of New York, Electricity. . .
9. Steven Ebbin and Raphael Kasper, Citizen Groups & the Nuclear Power
Controversy; Uses of Scientific & Technological Information (Cambridge,
Mass.: MIT Press, 1974), p. 36.
10. Association of the Bar of the City of New York, Electricity. . ., p. 69.
11. Arthur W. Murphy, "The National Environmental Policy Act and the Li-
censing Process: Environmental List Magna Carta or Agency Coup de
Grace?," Columbia Law Review, Vol. 72, October 1972, pp. 962-1007.
12. Ebbin and Kasper, Citizen Groups. . ., p. 30.
13. Calvert Cliffs' Coordinating Committee, Inc. v. Atomic Energy Commission,
449 F.2d 1109 (D.C. Cir. 1971).
175
-------
14. Calvert Cliffs' Coordinating Committee v. Atomic Energy Commission at
1122.
15. Ebbin and Kasper, Citizen Groups. . ., p. 51.
16. R. Gannon, "What are the Dangers?" The American City, Vol. 71, (January
1972).
17. U.S. Code, Title 10, Part 50, Section 109, 1972, p. 289.
18. Ebbin and Kasper, Citizen Groups. . ., p. 55.
19. U.S. Code, Title 18, Chapter 39, "Explosives and Other Dangerous Arti-
cles." (Washington, D.C.: U.S. Government Printing Office), 1977,
pp. 831-836.
20. U.S. Code, Title 46, Chapter 7, "Carriage of Explosives or Dangerous Sub-
stances," Section 170 (Washington, D.C.: GPO, 1971), pp. 11304-11310.
21. U.S. Congress, Senate. Clean Air Amendments of 1970, Public Law 91-604,
84 Stat. 1676. U.S. Code, Congressional & Administrative News, 91st
Congress - Second Session, 1970, Vol. 1 (Laws), pp. 1965-66.
22. U.S. Environmental Protection Agency, Office of Radiation Programs,
40 CFR 190, Environmental Radiation Protection Requirements for Normal
Operations of Activities in the Uranium Fuel Cycle, Final Environmental
Statement. Vols. I and II (EPA 520/4-76-016), November 1, 1976.
23. Reorganization Plan No. 3 of 1970. U.S. Code, Congressional & Adminis-
trative News, 91st Congress - Second Session, 1970, Vol. 3, pp. 6322-
6336.
24. Reorganization Plan No. 3 of 1970. . .
25. U.S. Environmental Protection Agency, 40 CFR 190, Environmental Radi-
ation Protection Requirements. . ., Vol. I, p. 7.
26. Ibid., Vol. I, p. 9.
27. National Academy of Sciences-National Research Council, The Effects on
Populations of Exposure to Low Levels of Ionizing Radiation, Report of
the Advisory Committee on the Biological Effects of Ionizing Radiation
(Washington, D.C.: NAS-NRC, January 1972).
28. U.S. Environmental Protection Agency, 40 CFR 190, Environmental Radi-
ation Protection Requirement. . ., Vol. I, p. 11.
29. Ibid., Vol. I, p. 9.
30. Ibid., Vol. I. p. 25.
31. Ibid., Vol. I, p. 26.
176
-------
32. Ibid. , Vol. I, p. 12.
33. Ibid., Vol. I, p. 26.
34. Ibid., Vol. I, p. 15.
35. Ibid., Vol. I, p. 21.
36. Ibid.
37. Ibid., Vol. I, p. 47.
38. Ibid.
39. Ibid., Vol. I, p. 48.
40. Ibid., Vol. I, pp. 80-86.
41. Ibid., Vol. I, pp. 50-51.
42. Ibid., Vol. II, p. 39.
43. Ibid., Vol. II, p. 21.
44. The discount rate implicit in the summation of lives is the solution to
the equation:
t=100 t-«
E V L = E V L / (1+rr
t=l C C t=l
Here, V is the valuation of life in the t-th year, L the number of lives
in that year, and _r is the discount rate, assumed to be constant. If the
number of lives per year were constant, out to infinity, and the valua-
tion of life a constant, the implicit interest rate would be 0.010101...
For the case at hand, there are more lives at risk in the earlier years,
and so the implicit interest rate is even lower. On the other hand, if
we assume an increasing absolute valuation of life, the discount rate
would be higher. Given the probable distribution of lives saved, it is
highly improbable that the implicit discount rate is higher than the
discount rate used for reducing control costs to present worths.
45. U.S. Environmental Protection Agency, 40 CFR 190, Environmental Radiation
Protection Requirements. . ._, Vol. I, p. 39.
46. Ibid., Vol. I, p. 49.
47. Ibid., Vol. I, p. 48.
48. Ibid.
177
-------
.15. Publicly Voiced Criteria for Nuclear Waste Material
Issues of how and where the United States should dispose of its nuclear
wastes have not focused in the mind of the average American. When asked about
possible dangers of nuclear power, only a small minority mention the waste
disposal problem. Yet when asked directly whether disposal of wastes repre-
sents a major problem that must be surmounted, a sizeable majority agree. En-
vironmental pressure groups rank the disposal problem high among risks of
nuclear power, but even the more sophisticated do not offer positive criteria
for an acceptable disposal alternative.
When the issue of disposal appears in nuclear power plant siting hearings
or other public forums, the principal response from environmental groups and
other citizens' bodies has been criticism of publicly available studies, and
an insistent reliance on process. Demands are typically made for further
studies to discover a "safe" disposal alternative. A milder response is to
request that evidence be presented to the public at large, in the apparent
hope that full and open discussion will in itself bring forth a generally ac-
ceptable disposal alternative.
Some critics of nuclear power development regard the issue of waste dis-
posal as subordinate, feeling it to be less dangerous and relatively unimpor-
tant compared with hazards of power plant operation. An example of this atti-
tude can be found in an article generally skeptical of the promises of safety
of nuclear power by Richard Rhodes. He says "... but the problem of nuclear
wastes disposal will almost certainly be solved." I/ By January, 1977, how-
ever, the perceived dangers of waste disposal were sufficiently newsworthy for
the National Broadcasting Company to show an hour-long program on the issue,
emphasizing the insidious dangers of unseen horrors. 2J
Even proponents of nuclear power can become opponents of waste storage if
it appears that the materials might be emplaced in close geographical proxim-
ity to themselves. In a letter commenting on the draft EPA Radiation Protec-
tion Requirements, two Iowa scientists thought the proposed standards were
"unscientifically based," "extremely short-sighted," "academic," and "inde-
fensible." They further commented that since "Iowa is a vital food production
State instrumental in feeding the nation and the world, it is hoped long-time
storage or processing of radioactive wastes in our State would be discour-
aged." 3/
Position of Public Interest Groups
Critical Mass, a monthly publication by the Ralph Nader-sponsored organi-
zation, lists reports in almost every issue from critics of ERDA's plans for
179
-------
waste management. These articles contain much information about the present
and future dangers of radioactive wastes. They contain many examples of what
critics call mismanagement of the EKDA search for disposal alternatives, but
there is so little in the way of what is considered a safe acceptable alter-
native that it is futile to use these critical reports in a search for posi-
tive criteria. Four principal themes are contained in the substance of these
criticisms:
1. Nuclear wastes are extraordinarily dangerous, representing
risks of a qualitatively higher order than any other ever
faced by society.
2. The Federal Government, and ERDA in particular, are guilty
of mismanagement, for too few resources have been directed
toward resolution of this problem. Continuation of nuclear
power plant development should be delayed until solutions
have been found to all possible dangers, including disposal
of wastes.
3. The government and private industry concerned with nuclear
power have hidden the truth about many dangers and problems
that remain to be solved.
4. Much uncertainty remains about certain basic scientific
facts that must be understood before any proposed disposal
alternatives can be judged as safe.
These attacks are not frivolous. Nor is it irresponsible to criticize
the presently offered disposal alternatives and point out management errors
without offering positive suggestions in such a highly technical policy field.
As noted, however, review of these criticisms sheds no light on what the cri-
tics would accept as a safe disposal alternative. They do not voice positive
criteria.
The most informed position among the public interest groups is that of
the Natural Resources Defense Council (NRDC). Their Citizens' Guide is far
superior to other public interest group reports in both quality and apparent
objectivity of handling of the issues. 4/
Their principal position is one of forcing the government to study fur-
ther the consequences of alternatives for disposal (in the generic sense of
disposal). They place reliance on further study by the government rather than
offering their own specific criteria for acceptability. They also urge citi-
zens to take an active part in NRG licensing procedures, and to intervene in
hearings for power plants and spent fuel reprocessing plants.
The NRDC sees disposal through transmutation as a promising option.
The elimination of 99.5 to 99.9% of the actinides in high-
level wastes through partitioning at reprocessing plants
and transmutation in existing light water reactors has been
shown to be a technically and economically feasible process.
180
-------
Higher efficiencies of removal seem likely in the near
future. However, making this a commercially practicable
procedure will require several years of government-financed
research and development. 5j
This position of seeing hope for transmutation remained alive in 1977.
In the opinion of Terry Lash, an author in Citizens' Guide, further instal-
lation of nuclear power should be stopped until increased funding for trans-
mutation research developed an acceptable disposal solution. 6J In Lash's
prepared statement before the Energy Facility Siting Council of Oregon, he
also suggested a specific rule to be applied to possible certification by the
state of proposed nuclear power plants.
A site certificate shall not be issued unless (a) there is
reasonable assurance of a thirty-year fuel supply for the
proposed nuclear power plant; (b) there is a fully approved
and licensed permanent disposal facility for commercial
high-level radioactive wastes with sufficient capacity to
contain thirty years' generation of high-level radioactive
wastes from the proposed nuclear power plant; and (c) the
site certificate applicant has obtained a firm, fixed-
price contract either for reprocessing of thirty years'
generation of irradiated fuel, including the disposal of
high-level radioactive wastes, or for the disposal of thirty
years' generation of irradiated fuel. ]_/
The first clause suggests that Oregon should not permit power plant sit-
ing until there is a reasonable prospect that power plant operations, once
begun, will not be interrupted or discontinued because of unforseen difficul-
ties related to the delivery of fuel. The second clause would delay issuance
of a siting certificate until a permanent disposal facility is actually in
operation. In effect, it regards the present-day practice of temporary stor-
age as unacceptable. Holding in abeyance the controversies about dealing on
a permanent basis with nuclear wastes, the rule would halt power plant devel-
opment and further production of wastes until a permanent, acceptable disposal
method is found. In his testimony, Lash also states:
. . .there is great uncertainty about whether or not the
federal government will correct existing inadequacies
of its program to meet its legal obligations and re-
solve the political difficulties associated with waste
disposal. _8/
In other words, no acceptable disposal alternative may ever exist, and
Lash suggests that Oregon delay power plant siting until the uncertainty is
resolved. His third clause suggests a practical test of whether a disposal
alternative really exists. Given the present ERDA policy of private operation
of waste reprocessing operations, adoption of this proposed rule would provide
an operational criterion of whether a disposal alternative existed, at least
within the contemplated 30 year life span of the suggested commitment.
181
-------
Lash's suggested rule comes as close as any offering in the public liter-
ature toward establishing a positive criterion for an acceptable waste manage-
ment alternative. His criterion is economic, not technological. Uncertainty,
to the Oregon siting board, which is the regulator of the power plants in
question, and the energy-consuming Oregon public, would be reduced along one
dimension. Economic uncertainty would be shifted to whatever body would
offer the "firm, fixed-price contract" for waste disposal.
Results of Public Surveys
Results of a survey by Louis Harris and Associates of public attitudes
toward development of nuclear power showed that perceived risks of nuclear
waste ranked higher than any other problem associated with nuclear power
development. _£/ The survey was conducted in March and April, 1975, with a
nationally representative sample of 1,537 persons aged 18 and over. In addi-
tion, 301 interviews were conducted with neighbors of nuclear power plants,
and 201 interviews with "leaders" — elected and appointed government offi-
cials, vice presidents of public affairs in large and medium companies, regu-
latory officials of agencies such as ERDA, NRG, EPA, and state commissions,
and members of environmental groups such as the Sierra Club.
In open-ended questioning on "the two or three main disadvantages of
nuclear power plants to produce electric power," 10 percent of the general
public mentioned problems with nuclear waste disposal, slightly more than the
9 percent of the presumably more sensitized nuclear power plant neighbors.
When asked the same question, nuclear waste was identified by 50 percent of
the environmentalists, 37 percent of the political leaders, 34 percent of the
regulatory officials, and 22 percent of the business leaders. While in this
open-ended format of questioning, waste disposal ranked fifth in the frequency
of response among the general public; it ranked first among environmentalists,
and regulators, and second among political leaders. Other problems which were
presented more often by the general public were generalized fears of power
plants being "unsafe, (and) dangerous," mentioned by 23 percent; and more spe-
cific risks such as the "danger of radiation contamination, leaks, cracks in
reactor," noted by 20 percent of the general public. The "danger of accii-
dents, explosions, and earthquakes was mentioned by 14 percent as shown in
Tables 10 and 11.
Under direct questioning, the dangers of wastes became the most critical
problem associated with nuclear power. When asked whether "the disposal of
radioactive waste materials which remain radioactive for many centuries to
come" was a major problem, 63 percent of the general public agreed, a greater
percentage than was associated with any other problems in the survey. "The
escape of radioactivity into the atmosphere," presumably the major danger as-
sociated with power plant operations, was thought to be a major problem by 49
percent. Among the leadership groups, direct questioning again brought wide-
spread agreement that waste disposal was a major problem.
In the open'-ended question format, risks of radioactive waste disposal
are mentioned by a relatively small percentage of the general public. How-
ever, when a direct probe brings forth the issue, a much higher percentage not
182
-------
TABLE 10
Extract of Results of Open-Ended Questioning on
"The Two or Three Main Disadvantages of Nuclear Power Plants
To Produce Electric Power"
(Percent Responding)
Total Political Business Regu- Environmen-
Disadvantage Cited Public Leaders Leaders lators talists
Unsafe, Dangerous 23
Danger of Radiation
Contamination, Leaks,
Cracks in Reactor 20
Danger of Accidents
Explosions, Earth-
quakes 14
Thermal Pollution
Kills Marine Life 12
Problems with
Radioactive Waste
Disposal 10
16
41
14
37
24
25
14
10
22
15
30
15
21
34
15
37
27
12
50
Source: Louis Harris and Associates, Inc., A_Survey of Public and Leadership
Attitudes Toward Nuclear Power Development in Llic United States, con-
ducted for Ebasco Services, Inc., New York, August 1975, page 55.
183
-------
TABLE 1.1
Agreement with Stated Problems Connected With Nuclear Power Plant?
(Percent Agreement)
Statement
Total
Public
Poll tical
Leaders
Business
Leaders
Regu-
lators
F.nvironmen-
talists
The Disposal of
Radioactive Waste
Materials Much
Remain Radioactive
for Many Centuries
to Come
Major Problem 63
Minor Problem 14
82
10
57
29
76
20
98
2
The F.scapc of
Radioactivity Into
the Atmosphere
Major Problem
Minor Problem
49
19
33
29
22
14
39
28
62
25
The Chance of an
Explosion in the
Case of an Accident
Major Problem
Minor Problem
47
28
14
37
18
29
17
43
45
29
The Discharge of
Warm Water into
Lakes and Rivers
that Could Endanger
Fish and Other
Water Life
Major Problem
Minor Problem
47
28
43
45
22
51
44
40
58
38
Source: Louis Harris and Associates, Inc., A Survey of Public and Leadership
Attitudes Toward Nuclear Power Development in the United States, con-
ducted for Ebasco Services, Inc., New York, August 1975, page 56.
184
-------
only thinks disposal a major problem, but the issue also tops the list of all
perceived problems. Two comments explain this apparent inconsistency. First,
the public may not perceive waste disposal as a systems problem of the entire
fuel cycle. Second, the problem may still be too sophisticated to be gener-
ally perceived by the public. The high number of interviewees who perceive
the risk as major may be a matter of the emotive language used — e.g.,
"remain radioactive for many centuries to come." On the other hand, "the pos-
sibility that plutonium, which is made in a nuclear power plant, could be
stolen by radical revolutionaries" (emphasis added) also contains emotive
language, yet was well down the scale of agreement as a major problem.
Another survey, conducted by the Battelle Memorial Institute, attempted
to probe more deeply into public attitudes and values toward waste disposal by
use of a highly structured questionnaire in conjunction with a variety of
sophisticated psychometric techniques. 10_/ The survey was conducted in early
1976, and obtained 465 usable responses from persons associated with a variety
of groups — nuclear technologists, environmentalists, high school and univer-
sity students, public utility employees, and members of civic organizations.
The basic purpose of the study was to probe and attempt to quantify four basic
dimensions of public attitudes toward nuclear waste disposal.
1. Short-term safety: those risks involved in the storage, trans-
portation, and emplacement of nuclear waste materials.
2. Long-term safety: that portion of total risk which would be-
gin after wastes were finally emplaced or disposed and which
would continue for the next 250,000 years.
3. Cost: the dollar cost required for a given waste disposal
method.
4. Accident detection and recovery: steps that could be taken to
reduce the consequences of an accident if it should occur after
final disposal. 11/
Despite the many elaborate psychometric techniques employed to analyze
the survey responses, the principal conclusions of the study are simple, oc-
casioning no surprises. By a significant margin, cost is felt to be the least
important factor associated with waste disposal methods. Among the other
three, Long-term safety received the highest priority by a small margin over
the other two factors.
Various scaling techniques were employed to attempt to quantify the priori-
ties. All methods seem suspect because no real trade-offs were ever necessary
— the answers may well have been heavily dependent on the particular scales
of the choices offered. Not surprisingly, the Nuclear Technologists rated
various levels of Short-term and Long-term risk as more satisfactory than did
other groups, while the Environmentalists gave the lowest ratings to various
levels of risk. In general, these two groups bounded the responses to all
attitudinal questions, with the other groups falling somewhere in between.
185
-------
Despite the very caveats that may be raised against the scaling tech-
niques employed, results of the study in this area should be reported. On
Short-term safety, Nuclear Technologists and Environmentalists were
the least and most risk-adverse, respectively, while views of the other groups
more closely resembled those of the Environmentalists. Reasonable levels of
satisfaction were associated with one death per year.
On Long-term safety, the mid-point on the scale again was associated with
one death per year with more extreme views held by Environmentalists and
Nuclear Technologists.
On the issue of Cost, feelings about Satisfaction were associated with a
scale running from $15.00 and higher per month in increased monthly costs of
electrical power. Given the low priority attached to Cost, even an increase
of $15.00 per month was near the mid-point of the scale, far from Very Unsat-
isfactory. General agreement was reached that increases of $3.00 to $5.00 per
month would be quite satisfactory.
The best source of views of the general public when actually confronted
with the possibility of having nuclear wastes stored in close proximity to
their places of work and homes comes from hearings on nuclear waste disposal
in Michigan, held by Representative Bob Carr, Democrat of East Lansing, pre-
siding. 12/ The hearings concerned ERDA's plan to conduct test drillings to
investigate suitability of the Salina salt beds in Michigan as possible sites
for nuclear waste storage. Statements were obtained from 26 persons, princi-
pally from local citizens' groups, environmental interest groups, and various
levels of Michigan government. The statements featured recitations about dan-
gers of transportation and possibilities of danger from nuclear wastes. The
unfortunate history of the Lyons, Kansas exploration was noted, along with
other tales of various mistakes made by the AEC and other bodies in charge of
nuclear wastes.
Much of the testimony was devoted not so much to staking out a position
either in favor of or against local disposal of nuclear wastes, but rather a
call for citizen input, full dissemination of information, and clear exposi-
tion of the processes and risks so that an informed citizenry could make a
decision. Perhaps the clearest exposition of this view was given in the pre-
pared statement of Congressman Philip E. Ruppe, Republican, of Houghton:
These decisions will have to be reached by all of us.
They cannot successfully be concluded by Congress
alone, by executive agencies, such as ERDA, alone, or
by state and local governments alone. We will all
have to develop that policy together, for we all have
a vested interest in the outcome of the decision.
As important as that final decision is, however, of
equal importance is the method we use to reach the
decision. It is my belief that there is only one path
that will make the ultimate choice a fruitful one.
That method involves open discussion and open debate
186
-------
about all alternatives involving every issue of the
nuclear power question. J_3/
The Governor of Michigan felt "strongly that information regarding pre-
liminary testing and future development of a proposed area must be made avail-
able to everyone." JL4/ This emphasis on process as necessary and correct in
order to reach the best decision was echoed by the Michigan United Conserva-
tion Clubs. After expressing that they v-ere "skeptical whether ERDA can prove
to the satisfaction of the state that disposal of nuclear wastes will be
safe," they stated:
"We believe the citizens of Michigan will agree with a
decision to site a nuclear dump in our State if the pro-
posal is fully aired and shown to be safe as supported
by the evidence." 15/
Summary
From the discussion above, there appears to be relatively limited concern
about risks related to nuclear wastes as compared to other risks within the
nuclear power cycle. Conversely, those who do perceive risks from nuclear
waste management are quite vocal and hold strong opinions often based on mis-
information, lack of knowledge, or both. Perhaps most important, there exists
a somewhat definitive, unanimous, and strong desire on the part of the public
for complete and absolute public disclosure of all facts related to nuclear
waste management decisions prior to the making of any decision. Operational
implication of these attitudes are quite clear: any risk analysis of nuclear
waste undertaken as part of the management process must be as detailed as
possible and must be made available to the greatest extent possible to the
general public.
Notes to Chapter 15^
1. Richard Rhodes, "Delusions of Power," The Benefits, Costs, & Risks of
Nuclear Energy. The Atlantic (June 1976), p. 41. In this report, the
views of two prominent environmental groups are summarized. The princi-
pal results of two surveys of attitudes towards nuclear waste disposal
are also presented. Useful evidence on public attitudes is contained
in Congressional hearings in Michigan relating to choice of a nuclear
waste disposal site in that state.
2. National Broadcasting Company, "Danger: Radioactive Waste," (January 26,
1977). Unfortunately, NBC did not collect audience participation statis-
tics on this program. We have no way of knowing whether this program
attracted more than the usual small audience for special news topics.
3. Rolf M. A. Hahne, and Robert L. Morris, to the EPA, reprinted in U.S.
Environmental Protection Agency, Office of Radiation Programs, 40 CFR
190: Environmental Radiation Protection Requirements for Normal Opera-
187
-------
tions of Activities in the Uranium Fuel Cycle, Final Environmental State-
ment (EPA 520/4-76-016), Vol. II, p. A. 14-15.
4. Natural Resources Defense Council, Inc., Citizens' Guide: The National
Debate on the Handling of Radioactive Wastes from Nuclear Power Plants,
by Terry R. Lash, John E. Bryson, and Richard Cotton (Palo Alto, Natural
Resources Defense Council, Inc., November 1975).
5. Natural Resources Defense Council, Inc., Citizens' Guide. . . pp. 25-26.
6. Telephone conversation with Terry Lash, Natural Resources Defense Coun-
cil, February 2, 1977.
7. Terry R. Lash, "Prepared Testimony Before the Energy Facility Siting
Council of Oregon," January 11, 1977 (mimeo) p. 5.
8. Terry R. Lash, "Prepared Testimony Before the Energy. . .," p. 16.
9. Louis Harris and Associates, Inc., A Survey of Public and Leadership
Attitudes Toward Nuclear Power Development in the United States, conduc-
ted for Ebasco Services, Inc., New York, August 1975.
10. Battelle Memorial Institute, Human Affairs Research Center, Public Values^
Associated with Nuclear Waste Disposal, by William S. Maynard et al.
BNWL-1997, UC-70 (Richland, Wash.: Battelle, Pacific Northwest Labora-
tories, June, 1976).
11. Battelle Memorial Institute, Public Values Associated with Nuclear Waste
Disposal. . ., p. 9.
12. U.S. Congress, House, Committee on Interior and Insular Affairs, Subcom-
mittee on Energy and the Environment, Nuclear Waste Disposal in Michigan,
Oversight Hearing before the Subcommittee on Energy and the Environment,
94th Congress, 2nd Session, 1976.
13. U.S. Congress, House, Nuclear Waste Disposal in Michigan. . ., p. 7.
14. U.S. Congress, House, Nuclear Waste Disposal in Michigan. . ., p. 11.
15. U.S. Congress, House, Nuclear Waste Disposal in Michigan. . ., p. 143.
188
-------
16. Evaluation of the Decision Process as Applied
to Nuclear Waste Management
Options for managing nuclear wastes are numerous, as discussed previous-
ly. Management of nuclear waste can be meaningful only if the public is as-
sured that risks are minimal and if cost-benefit analysis indicates a socially
acceptable outcome. The option selected must be virtually flawless in order
to gain credibility from the general public as well as the scientific commun-
ity. At the present time, few members of the scientific community and perhaps
even the general public favor disposal concepts such as extraterrestrial
transport. Consequently, geological isolation appears to be the only current-
ly viable alternative. Thus, appropriate risk analysis of this disposal con-
cept is an integral part of the decision process. Risk analysis can essen-
tially be based on either of two approaches: (1) evaluation on the basis of
empirical data (an actuarial approach); or (2) evaluation of risks based on
predictive techniques which combine known subset failure data with an under-
standing of system failure modes.
The actuarial approach is appropriate where data from similar past occur-
rences are available and the future is expected to be similar to the past.
For example, reasonably good data exists on transportation accidents. These
data can be used to predict accident frequencies involved in high-level waste
transport, provided that the transport is effected by conventional means.
However, no complete set of data exists on the reliability of high-level waste
storage facilities. Only partial data are available for assessing the risk
associated with storing high-level wastes. Therefore, for these facilities,
risk analysis such as the fault tree methodology used in the Rasmussen study
to predict reactor safety must be used in spite of data problems associated
with such an analysis.
Risk Analysis of Geologic Disposal Methods
The principal difficulty with data related to geological disposal of
nuclear wastes arises from the instability of geological formations over long
periods of time. A first approach would be to estimate future risks on the
basis of past performance. Thus, in assessing future stability against earth-
quakes, volcanoes, or the presence of water, the geologic age since the last
occurrence of any of these events might be used as an estimate of the future
time scale over which such an event would not occur again. The assumption of
this homogeneity in nature, however, could be misleading. Over periods con-
sidered short in comparison to the geologic times involved, this approach is
probably acceptable. For example, in the near future, areas of low earthquake
189
-------
probability in the past will be subject to fewer earthquakes than areas of
past high earthquake frequency. For predictions over long periods of time,
however, past performance may be a poor indicator of future performance as the
earth is in a state of noticeable dynamic change. For example, present plate
tectonic theory shows that the ocean floor is in a constant state of change,
with new material being added continuously, drifting across the ocean, and
then being subducted below the earth's crust again. In light of this phenome-
non, the oldest rocks will have the shortest future life while newly created
rocks will have the longest future life.
These problems are fundamental in analysis of risks associated with geo-
logical disposal of nuclear waste. Some of these problems, however, can be
minimized by appropriate risk analysis involving six basic tasks:
1. Estimation of engineering and component failures
2. Estimation of human failures
3. Estimation of unanticipated natural events
4. Pathway analysis
5. Receptor distributions analysis
6. Health effects analysis
A brief description of these tasks is provided below.
Engineering and component failures. For the geological disposal concept,
reliance on engineering is relatively large. However, this task is not an
overriding component of a hazard analysis of waste management, because most of
the required data are available and research methodology well established.
Human failures^. Human failures can be estimated fairly accurately in the
limited sense of routine mistakes during known operating steps. However,
estimation of human failures in the larger sense of social or political insta-
bility is clearly uncertain. This uncertainty has a high impact on waste dis-
posal concepts which require continued maintenance or surveillance by man,
such as storage in engineered surface units. For systems with relatively
short lifetimes, the inability to estimate future social and political stabil-
ity may not be critical, but for systems with anticipated lifetimes on the
order of a hundred or more years, the inability to analyze the risk from
large-scale human failure makes it difficult to assure future generations that
small risk is involved. This fact is a definite disadvantage of engineered
storage systems which might require human intervention to maintain integrity.
Another area with a high degree of uncertainty is the estimation of fail-
ure from deliberate human action such as sabotage. Additional research is
needed before a low risk from storage in areas easily accessible to man, such
as surface storage, could be assured.
190
-------
Unanticipated natural events. The estimate of natural disasters through
relatively frequent occurrences such as tornadoes or lightning strikes is well
understood. The risk from events such as earthquakes over the short term is
also reasonably well known, but is less understood over the longer-term. The
risk from longer-term geologic events such as ice ages, melting of the ice
caps, and change of continental shapes is relatively unknown. The risk from
dependence upon geologic stability over short periods of time is fairly well
known, but the risk associated with dependence upon geologic stability over
geologic times (on the order of a million years) is essentially unknown.
These facts tend to rule against disposal schemes in which isolation is needed
for long periods and the disposal medium is either relatively vulnerable or
does not provide good isolation from biota. This would include disposal in
vulnerable salt deposits or disposal in sea or ice environments which could
lead to rapid dispersal of the waste material to living organisms. If long-
term isolation of wastes is needed, deep rock formations which provide good
isolation and invulnerability characteristics as well as apparent stability
would be a preferable medium.
Pathway analysis. Uncertainties in pathways analysis do not appear as
significant as uncertainties in estimating the initial basic failure. Limit-
ing consequences can always be calculated by making conservative pathways
assumptions. However, basic data on the travel of radionuclides over long
time scales are only partially understood. The migration of many different
isotopes through soil and concentrations in various media have been the sub-
ject of numerous studies. These data typically encompass a few years or per-
haps tens of years' data, and over-reliance on holdup in soil should not be
relied upon for extended periods of time. This is significant in cases such
as Hanford, where retention in the soil is presumed to isolate large quanti-
ties of long lived isotopes from more critical systems. Continued reliance
upon such mechanisms, or future reliance on such mechanisms in other disposal
schemes, does not appear wise given the long-term pathway uncertainties.
Receptor distributions. Receptor distributions and ultimate health ef-
fects are fairly well known, and only large relative uncertainties are likely
to have profound effects on the choice of disposal methods. However, some un-
certainties of this nature do exist. The most obvious uncertainty pertains to
the health effect of long lived alpha emitters, such as plutonium.
Health effects. Consideration of health effects in risk analysis manage-
ment is always paramount, and yet ultimately these effects are relatively un-
known. Large-scale leakage would almost always be catastrophic. Slow unde-
tected leakage could have global effects half a century from today. In the
same manner that very fine particulates take from 20 to 40 years to produce
cancer, suspicions that radioactivity and its effects on humankind might be-
have in a similar way seem realistic. The effects of such radiation over the
long term are much more subtle, but could well be far reaching and more de-
structive for much longer periods of time.
If quantitative risks can be determined for different waste management
concepts, one possible way to choose among the concepts would be to minimize a
summation over all types of release events integrated over all time; that is,
minimize the quantity
191
-------
Ei ft pi(t) ci(t) Vt)dt
where P.(t) is the time dependent probability of release for release type i,
C^(t) is the consequence of this release, and T.(t) is a future weighting fac-
tor.
Regarding the choice of a future weighting factor, it might be possible
to assign equal weight to future risks as present risks; i.e., let T. equal
unity for each future time period. Another approach would be to discount fu-
ture risks in the same manner that future costs are discounted in making eco-
nomic comparisons. The choice of a future weighting factor depends upon our
perceptions of the future world and our responsibility for the future. The
choice is clearly societal, not technological.
With regard to future weighting factors, the type of risk posed must be
considered. If it is known that future risks can be modified with a reason-
able amount of capital or effort, these future costs should probably be dis-
counted in making present decisions. However, in the case of nuclear waste
release, the likely risk will often be an irretrievable loss of land areas,
water systems and human activity for very long periods of time. The principal
difficulty with these "future weighting factor(s)" is, of course, the need to
relate them to the value of the consequences that may result. Establishing
values acceptable to all major groups in society can be a difficult if not
impossible undertaking.
Health risks are obviously a critical area for consideration. Over the
short run, a future cancer risk to the population is probably of equal nega-
tive value as a present cancer risk. However, over very long periods of time,
future and present cancer risks are not comparable if cures for cancer can be
assumed over the long run. In that case, greatly distant cancer risks should
be largely discounted in comparison to present cancer risks.
Another problem in the use of overall risk summations as a measure by
which alternatives can be evaluated is that the same average number of fatali-
ties may be incommensurate. This is especially true when considering time-
integrated total risks.
Waste Management Decisions
Clearly, the overriding issue is whether nuclear waste management poses
an acceptable risk, or whether nuclear power should be curtailed to avoid
creating further nuclear waste. This decision is not simply technological.
Specifically, societal perceptions of future risks are an important part of
any waste management decision. However, technological analysis suggests that
if nuclear wastes are carefully handled (for instance, fractionation recycle
and transmutation of long-lived activity), the potential risks may be no
greater than those posed by many of man's presently accepted activities, even
accounting for the full future risk potential of nuclear waste. Mishandled
192
-------
nuclear waste, on the other hand, could pose a truly significant hazard to
this generation and many future generations.
At the level of selecting among waste management alternatives, certain
features stand out. First, the cost of most waste disposal systems represents
only a small fraction of the total cost of electricity. At the same time,
mistakes in waste handling can lead to large future cost commitments, or even
large potential health risks. This possibility suggests that any decisions in
this regard should undergo much more complete risk analysis than has been made
in the past.
Although costs of waste management are likely to be a small fraction
(typically less than 1%) of the total cost of electricity, they will represent
large absolute costs. !_/ Awareness of this likelihood is particularly import-
ant: when these costs may be borne entirely by an isolated segment of the total
nuclear electricity cycle. The degree of actinide separation in reprocessing
spent fuel provides a good example for potential mismanagement. Within the
present nuclear fuel cycle, the spent fuel reprocessor must operate on the
small savings represented by the recycle of spent fuel. 2j Regulations govern-
ing this part of the fuel cycle could conceivably result in no fuel proces-
sing at all. From the waste management standpoint, this would result in maxi-
mizing the amount of long-lived activity in high level waste. Waste manage-
ment decisions and regulations must be taken in the context of the entire
nuclear picture. Also, where society would benefit by requiring certain waste
management practices, methods should be considered to assure that the costs of
such decisions and regulations are borne by the ultimate users of electricity,
and not by an isolated portion of the nuclear fuel cycle.
Another good example of how risks from waste management affect risk in
other activities is the trade-off which must be made between immediate risks
posed by plutonium recycle and the long-term reduction in risk which can be
achieved by eliminating plutonium through recycle. These risks are interde-
pendent. To analyze, regulate or manage these risks in isolation could lead
to wrong social decisions, with serious consequences.
At present over 20 different agencies are involved in attempting to co-
ordinate nuclear power management decisions. These agencies are carrying out
supposedly well established goals, unique to each agency. However, the oper-
ation of nuclear plants, guidelines for management, and regulations for licen-
sing power plants all require a unique kind of coordination and exchange of
information if correct trade-offs and decisions are to be made. The difficul-
ties faced by these 20-odd agencies in trying to achieve coordinated efforts
appear at times overwhelming. It should not be surprising that the situation
often results in fragmentation and lack of communication.
Criteria for Evaluating the Decision Process
The problem of deciding upon the appropriate waste disposal alternative
depends on the low probability of a bad occurrence with any seriously proposed
waste disposal mechanism, coupled with the high consequence of an accident.
Limited knowledge concerning many aspects of nuclear waste disposal implies
193
-------
that complete agreement on the expected outcome is impossible. Thus, no
matter which alternative is selected, any decision now on the appropriate form
of nuclear waste disposal would generally elicit a significant divergence of
opinion.
In addition to the general limitations on knowledge, nuclear wastes hold
profound implications for future generations. Some recognition of preferences
regarding future generations must be made during the decision process. The
decision among waste disposal alternatives is important, yet the uncertainty
inherent in any waste disposal mechanism implies that the ex post correct de-
cision will not be known for many years. Public acceptance of the selected
disposal mechanism therefore requires that the process be one which the public
believes will give an incorrect decision a very low probability.
Part I of this study developed a set of criteria for evaluating the so-
cial decision process in situations such as nuclear waste disposal. In this
concluding Chapter of Part III, we shall evaluate the decision process used in
nuclear waste disposal by reference to the eight criteria specified in Part I,
Chapter 4. For ease of reference, these criteria are listed in the accompany-
ing Table 12.
1. Are parties who incur the present costs adequately represented?
Parties presently at risk are all members of the general public who may
be exposed to radiation from nuclear waste. Costs are borne by members of the
public who will suffer health damage as the result of this exposure. As in
the case of pesticides, some of the monetary part of this burden, namely the
cost of treatment covered by health insurance, is spread directly among the
general public. However, since everybody is potentially at risk, to quite a
large degree it is the general public that may be affected.
During the earlier stages in development of nuclear power, the AEC at-
tempted to limit the scope of public hearings on new facilities. Groups liv-
ing near the proposed sites were heard. However, the AEC tended to exclude
from consideration broad issues of public interest, such as the creation and
management of nuclear wastes. This attitude was one factor underlying the re-
organization of the government's nuclear programs. The interest of the gener-
al public is now represented by NRC, EPA and Department of Energy from within
the public sector. Private sector still has had only marginal representation
in the issue of nuclear waste disposal per se, because of its emphasis on
other issues related to the nuclear power cycle and because the final criteria
and related issues are still being developed by the Federal Government.
As noted in the case study on pesticides, the adequacy of representation
on behalf of parties who bear the present costs depends on the financing of
those representing the general public. Because nuclear waste management is a
complex and highly technical subject, the need for expertise is cardinal. At
the same time, most environmental action groups are poorly financed. If the
private sector is to be adequately represented, some means may have to be
found to enable the private sector to pay for this expertise. This raises the
issue of whether some means should be instituted for compensating parties who
intervene in nuclear waste management proposals on the side of those who bear
194
-------
TABLE 12
Criteria for Evaluating Decision Processes
Embodying Low Probability, High Consequence Risks
1. Are parties who incur the present costs adequately represented?
2. Are parties who receive the present benefits represented?
3. Are interests of future generations suitably considered?
4. Are the distributional aspects considered systematically?
5. Are the social decision-makers unbiased and open?
6. Is the development of knowledge concerning the effect of possible
alternative decisions adequate?
a. Do views that appear to come from reasonable sources receive
— or appear to receive — "equal time and treatment"?
b. Are all significant differences considering the quality of
knowledge included?
c. Are the gaps in knowledge identified?
d. Are procedures for collecting knowledge of the area efficient?
e. Are the plausible consequences spelled out for the worst
contingency?
f. Are alternatives suitably considered?
7. Is information given to the individuals who are at risk?
8. Can a decision change if significant new information is obtained?
it
Source: Part I, Theoretical Considerations, Chapter 4, pp. 31-35.
195
-------
the present (or future) costs. Precedent for such compensation exists within
the legal process. In a class action suit, the attorneys representing the
plaintiffs can expect to receive fees if they are successful. Further, expert
witnesses and related costs can also be expected to be paid in such suits.
One additional overriding factor favors such compensation in the case of
nuclear waste. We are referring, of course, to the fact that management of
nuclear waste is undertaken by the Federal Government and therefore there
exists the danger of a conflict of interests on the part of Government
institutions/agencies that regulate waste disposal. This factor does not
exist in the case of pesticides.
2. Are parties who receive the present benefits represented?
No single population group benefits directly from nuclear waste manage-
ment. Of course, those who benefit from nuclear power will gain measurably if
any scheme for management or disposal of nuclear wastes were to receive broad
social acceptance. Proponents of nuclear power can be expected to have ade-
quate representation at any hearing relating to nuclear waste.
3. Are the interests of future generations suitably considered?
Unlike the case of pesticides, potential impacts of nuclear waste tran-
scend generations. As a matter of fact, no other man-made event creates longer
lasting effects than waste resulting from nuclear power production. As indi-
cated in previous chapters, considering the interests of future generations is
one of the most difficult issues. As also indicated previously, this issue
may lie beyond the competence of economists or administrators and in the realm
of theologians and philosophers.
Little can be said here about this issue except to note that if waste
management uses all six task analyses described in the first part of this
chapter, the interests of future generations are considered as much as is
possible, given the unique characteristics of nuclear waste.
4. Are distributional aspects considered systematically?
To date, no party in the debate on nuclear waste disposal has raised dis-
tributional aspects as an important issue. That is, all parties have impli-
citly assumed that benefits and costs of disposal activities will be spread
widely among the population. This may well be the case. There is no estab-
lished mechanism, however, for systematically inquiring into distributional
aspects.
5. Are decision-makers "unbiased" and open?
A number of issues relate to this criterion. First, as described else-
where in this study, nuclear waste management does not have well-established
and prescribed procedures such as those which exist for pesticides. Second,
some of the procedural issues and certainly the final criteria are still being
considered. Current procedures will certainly be modified. Finally, Federal
196
-------
Government officials may harbor biases toward certain waste management options
with failure modes which are difficult to detect and even more difficult to
correct.
The fact that the government has not adopted or embarked on any "perma-
nent" disposal method can, on the one hand, be interpreted as displaying a
high degree of unbiasedness. On the other hand, the government continues to
accumulate increasing amounts of nuclear waste in "temporary" storage facili-
ties. This indicates an implicit bias — or faith — that an acceptable per-
manent solution is technologically feasible and will be determined before the
wastes held in temporary storage become unmanageable. At a minimum, the
decision process has resulted in a compromise position. Such a decision pro-
cess is not without bias. However, in view of the fact that development of
nuclear power has noticeably slowed, the state of affairs just described may
reflect bias in the former AEC decision process more than in the current
decision process.
6. Is development of knowledge concerning the effect of possible
alternative decisions adequate?
a. Do views that appear to come from reasonable sources receive —
or appear to receive — "equal time and treatment?"
Under the AEC it is not clear that reasonable views did in fact receive
"equal time and treatment." The AEC no longer exists, however, and waste man-
agement discussions are a continuing dialogue. Therefore it is difficult to
answer this question. With the continued growth of broad-based public inter-
est groups, on net balance it would appear that reasonable views do now re-
ceive equal time and treatment.
b. Are all significant differences considering the quality o^f
knowledge included?
Analysis of past experience in the nuclear waste management discussion
clearly indicates that there have been considerable differences regarding the
quality of knowledge pertaining to nuclear waste management. These differ-
ences have been brought out in the open and discussed in some detail, but it
is not clear whether or how these differences are taken into account in
"official" considerations.
c. Are the gaps in knowledge identified?
Analysis of the existing literature and documents on nuclear waste man-
agement identifies many gaps in knowledge and these have been discussed in
considerable detail. It is likely, however, that in some cases no additional
expenditures or effort will narrow these gaps within a short term period.
Thus, although in the case of nuclear waste management there is adequate in-
formation on the existence of gaps in knowledge, some of these continue to
exist due to the nature of the subject.
197
-------
d. Are procedures for collecting knowledge of the area efficient?
Analysis of past experiences in this area, presented in Chapters 13 and
14, indicates that considerable thought has been given to the collection of
knowledge pertaining to nuclear waste management. In some cases, however,
even the conceptual and theoretical underpinnings are in dispute. This makes
the process more difficult. A principal problem pointed up by this criterion
is that there are no well-established procedures for collecting and dissem-
inating some of the information required.
e. Are plausible consequences spelled out for the worst
contingency?
A number of documents spell out in some detail the consequences of a
number of contingencies. The potential disasters that could result from
nuclear waste are so great that they invite the development of "doomsday"
scenarios. The principal difficulty that arises from application of this
criterion to nuclear waste concerns which waste management concepts would
result in the worst contingency under certain conditions.
f. Are alternatives suitably considered?
As presented in Chapter 12, a number of alternatives have been considered
in considerable detail, and it appears that this criterion has been satisfied
by the waste management process.
7. Is^ information, given to individuals who are at risk?
As already described there now appears to be a deliberate effort by Fed-
eral Government to provide the public with information pertaining to nuclear
waste management, hence this criterion appears to be satisfied.
8. Can a decision change if significant new information is obtained?
This evaluation criterion, as much as any other, serves to protect the
interests of future generations. The issue raised by this evaluation criter-
ion is perhaps the greatest source of difficulty for all "permanent" geologi-
cal disposal methods that have been proposed to date. Some methods may be
more reversible than others. For example, wastes suspended in polar ice packs
could be dug up and retrieved more easily than wastes that have been perma-
nently fused into deep rock wells. Herein lies the rub. Waste disposal
methods considered best by certain other criteria tend to be the most com-
pletely irreversible once completed. It is considerations of this sort which
led the Natural Resources Defense Council (NRDC) to favor greater expenditures
on transmutation research. To the extent that this is a valid and important
evaluation criterion, it would appear that NRDC's priorities have considerable
merit.
198
-------
Summary and Conclusion
As discussed above, application of these eight evaluation criteria to
the decision process for selecting among nuclear waste management options is
often difficult because of certain unique characteristics of nuclear waste.
Each criterion is subject to dispute and possible refinement. However, they
do permit a broad-based evaluation of the decision process used to select
among waste disposal alternatives. They also assure protection of the rights
of individuals — now and in the future — affected by the decision. This
latter assurance should enhance public acceptability of the final decision.
Although public concern is often an issue in matters of policy decision,
seldom has the public been more concerned than in issues dealt with here. In-
difference to societal input is totally out of the question where issues of
nuclear power are concerned. At one extreme, the public could conceivably put
nuclear power out of business. Or they could delay action to the point where
energy requirements of the nation and world became even more critical than is
currently predicted.
Another implication of this case study, though briefly mentioned, is the
lack of a complete systems approach toward nuclear waste management by numer-
ous governmental entities now responsible for various aspects of nuclear
energy. There can be no substitute for communication and informed action by
these many governmental units. A comprehensive program plan for mutual ex-
change, as well as structured cooperation in their efforts, objectives and
goals is needed to prevent fragmentation of our national energy policy for
nuclear power.
Finally, this study indicates that nuclear waste disposal is not without
grave risks, but equally, is not beyond the point of safe control—both
technologically and managerially, now and in the future. The thoroughness of
the approach used to resolve problems of disposal is unparalleled in magni-
tude, and should be so perceived. However, further studies on risk analysis
methodologies and benefit-cost analysis should be carried out prior to selec-
tion of final criteria. These studies should indicate where conceptual data
do not currently exist. Given the time for such studies prior to committing
to a permanent method for nuclear waste disposal, the final decision can be
made with a very high probability of being correct.
Notes to Chapter 16
1. The building of 1000 nuclear plants represents a present worth capital
cost of about 500 billion dollars, which is equivalent to about 50
billion dollars on a levelized annualized annual basis. A one percent
waste management cost would represent 500 million dollars/year.
2. If reprocessing costs per unit of uranium and/or plutonium are higher
than the cost of an energy equivalent amount of newly mined uranium,
utilities will clearly have no incentive to purchase reprocessed fuel
material.
199
-------
References for Part III
AAS Report, "After 10 Years: Thyroid Damage," Science News, 93 (January 13,
1968).
Adkins, C. R. and C. G. Fencelet, "Nuclear Reactor Safety." Pittsburgh, Pa.:
School of Engineering, Carnegie Mellon University, 1974.
Alfven, Hannes, "Fission Energy and Other Sources of Energy," Science and
Public Affairs, The Bulletin of the Atomic Scientists, (January 1974)
pp. 4-8.
Allied Gulf Nuclear Services, Environmental Report for Barnwell Nuclear Fuel
Plant, November 5, 1971.
Amberson, C. B. and D. W. Rhodes, "Treatment of Intermediate- and Low- Level
Radioactive Wastes at the National Reactor Testing Station (NRTS),"
Meeting on Practices in the Treatment of Low- and Intermediate- Level
Radioactive Wastes, Vienna, International Atomic Energy Agency. (IAEA-
SM-71142. 1966) DP. 419-437.
Ambrosek, R. G., Materials Testing Reactor Operating Manual, Volume III
Contaminate Air Effluent and Water Disposal Systems, October 1967. (CI-
1064).
American Society of Mechanical Engineers. Boiler and Pressure Vessel Code.
"Section XI — Rules for In-Service Inspection of Nuclear Reactor
Coolant Systems." New York: ASME, January 1970.
"Antarctic Treaty," Department of State Bulletin (December 31, 1959).
Association of the Bar of the City of New York, Electricity and The Environ-
ment: The Reform of Legal Institutions. Special Committee on Electric
Power and the Environment, Sheldon Oliensis, Chairman. St. Paul, Minn.:
West Publishing Company, 1972.
Atomic Industrial Forum, Inc. — British Nuclear Forum. International Con-
ference, Royal Garden Hotel, London, September 10-13, 1974.
Bailar, J. C., III, and J. L. Young, Jr., "Oregon Malignancy Pattern and
Radioisotope Storage," Public Health Reports, 81 (April 1966) pp. 311-
317.
201
-------
Battelle Memorial Institute. Aquatic Bioenvironmental Studies in the
Columbia River at Hanford 1945-1971; a Bibliography with Abstracts.
Richland, Wash.: Battelle, Pacific Northwest Laboratories, 1973. (BNWL
1734).
Battelle Memorial Institute. Characterization of Actinide Bearing Soils:
Top Sixty Centimeters of 216-Z-9 Enclosed Trench, by L. I. Ames. Rich-
land, Wash.: Battelle, Pacific Northwest Laboratories, 1974. (BNWL
1812).
Battelle Memorial Institute. A Critical Review of Biological Accumulation,
Discrimination and Uptake of Radionuclides Important to Waste Management
Practices, 1943-1971, by K. R. Price. Richland, Wash.: Battelle,
Pacific Northwest Laboratories, 1971. (BNWL-B-148).
Battelle Memorial Institute. Disposal Characteristics of Plutonium and
Americium in a High Salt Acid Waste, by B. F. Hajek and K. C. Knoll.
Richland, Wash.: Battelle, Pacific Northwest Laboratories, 1966.
(BNWL-CC-649).
Battelle Memorial Institute. Effect of Hanford Plant Operations on the
Temperature of the Columbia River, 1964 to Present, by R. T. Jaske and
M. 0. Synoground. Richland, Wash.: Battelle, Pacific Northwest Labo-
ratories, 1970. (BNWL 1345).
Battelle Memorial Institute. Environmental Status of the Hanford Reserva-
tion, by P. J. Bramson et al. Richland, Wash.: Battelle, Pacific
Northwest Laboratories, September 1973. (BNWL-B-278).
Battelle Memorial Institute. Environmental Surveillance at Hanford for CY-
1970 to CY-1974. Richland, Wash.: Battelle, Pacific Northwest Labo-
ratories. (BNWL-1669, UC-41; -1683, UC-41; -1811, UC-41; -1910, UC-41).
Battelle Memorial Institute. Evaluation of Radiological Conditions in the
Vicinity of Hanford for 1964-1969. Richland, Wash.: Battelle, Pacific
Northwest Laboratories. (BNWL-90; -316; -439; -1341, UC-41; -1505,
UC-41) .
Battelle Memorial Institute. Feasibility Evaluation and R & D Program Plan
for Transuranic Partitioning of High-Level Fuel Reprocessing Waste, by
J. W. Bartlett et al. Richland, Wash.: Battelle, Pacific Northwest
Laboratories, November 1973. (USAEC Report BNWL-1776).
Battelle Memorial Institute. High Level Radioactive Waste Management Alter-
natives, Vol. I. Richland, Wash.: Battelle, Pacific Northwest Labora-
tories, May 1974. Vol. I (BNWL-1900).
Battelle Memorial Institute. Long-Term Study of Inhaled Plutonium in Dogs,
W. J. Blair et al. Richland, Wash.: Battelle, Pacific Northwest Labo-
ratories, 1966. (AFWL-TR65-214).
202
-------
Battelle Memorial Institute. Methodology for Risk Analysis of Nuclear Waste
Management System, by T. H. Smith, et al. Richland, Wash.: Battelle,
Pacific Northwest Laboratories, 1974. (BNWL-SA-4899') .
Battelle Memorial Institute. "Nuclear Waste Management and Transportation
Quarterly Progress Reports, September-December 1974 and January-March
1975," Task 5, Richland, Wash.: Battelle, Pacific Northwest Labora-
tories.
Battelle Memorial Institute. Human Affairs Research Center. Public Values
Associated with Nuclear Waste Disposal, by William S. Maynard, et al.
Richland, Wash.: Battelle, Pacific Northwest Laboratories, June 1976.
(BNWL-1966, US-70).
Battelle Memorial Institute. Quantitative Evaluation of Environmental
Factors Affecting Population Exposure Near Hanford, by J. F. Hornstead.
Richland, Wash.: Battelle, Pacific Northwest Laboratories, 1970.
(BNWL-SA-3203).
Battelle Memorial Institute. Quarterly Progress Report, Research and Devel-
opment Activities, Waste Fixation Program, July through November, 1972.
Richland, Wash.: Battelle, Pacific Northwest Laboratories, 1972.
(BNWL-1699) .
Battelle Memorial Institute. Radiological Evaluation for Advanced Waste
Management Studies, by D. H. Denham, et al. Richland, Wash.: Battelle,
Pacific Northwest Laboratories, September 1973. (BNWL-1764).
Battelle Memorial Institute. Radiological Status of the Groundwater beneath
the Hanford Project, July-December, 1972, by K. L. Kipp. Battelle,
Pacific Northwest Laboratories, August 1973. (BNWL-1752).
Battelle Memorial Institute. Radiological Status of the Groundwater beneath
the Hanford Reservation, January-December, 1973, by K. L. Kipp. Rich-
land, Wash.: Battelle, Pacific Northwest Laboratories, April 1975.
(BNWL-1860).
Battelle Memorial Institute. Reactions of Organic Wastes and Soils, by
K. C. Knoll. Richland, Wash.: Battelle, Pacific Northwest Labora-
tories, 1969. (BNWL-860).
Battelle Memorial Institute. "Research & Development Activities Waste Fixa-
tion Program. Quarterly Progress Report," by J. L. McElroy. Richland,
Wash.: Battelle, Pacific Northwest Laboratories, July 1974. (BNWL-
1841).
Battelle Memorial Institute. A Review of Studies on Soil-Waste Relation-
ships on the Hanford Reservation from 1944 to 1967, by R. C. Routson.
Richland, Wash.: Battelle, Pacific Northwest Laboratories, March 1973.
(BNWL-1464).
203
-------
Battelle Memorial Institute. The Risk of Transporting Plutonium Oxide and
Liquid Plutonium Nitrate by Truck. Richland, Wash.: Battelle, Pacific
Northwest Laboratories, December 1974. (BNWL-1846).
Battelle Memorial Institute. Transmutation of High-Level Radioactive Waste
with a Controlled Thermonuclear Reactor, by W. C. Wolkenhauer, et al.
Richland, Wash.: Battelle, Pacific Northwest Laboratories, September
1973. (BNWL-1772).
Battelle Memorial Institute. The Use of Hanford Waste Ponds by Waterfowl and
Other Birds, by R. E. Fitzner and K. R. Price. Richland, Wash.:
Battelle, Pacific Northwest Laboratories, 1973. (BNWL 1738).
Battelle Memorial Institute. Vascular Plants of Waste Storage Sites in the
200 Areas of the Hanford Reservation, by K. R. Price and W. H. Rickard.
Richland, Wash.: Battelle, Pacific Northwest Laboratories, 1973. (BNWL
1796).
Battelle Memorial Institute. Waste Solidification Program Summary Report,
Volume 11, Evaluation of WSEP High-Level Waste Solidification Processes,
by R. L. McElroy, et al. Richland, Wash.: Battelle, Pacific Northwest
Laboratories, July 1972. (BNWL-1667).
Bendixsen, C. L. and G. E. Lohse, Storage Facilities for Radioactive Calcined
Waste Solids at the Idaho Chemical Processing Plant. Idaho Falls, Idaho:
Idaho Nuclear Corporation, July 1968. (NTIS Ordering No. IN 1155).
Bendixsen, C. L., G. E. Lohse, and M. P. Hales, The Third Processing Campaign
in the Waste Calcining Facility. Idaho Falls, Idaho: Idaho Nuclear
Corporation, May 1971. (NTIS Ordering No. IN 1474).
Bendixsen, C. L., and K. L. Rohde, Proposed Demonstration Program for
Recovery and Disposal of Fission Product Krypton and Xenon at the Idaho
Chemical Processing Plant. April 1972. (ACI 119).
Bergstrotn, Stig 0. W. , "Environmental Consequences from the Normal Operation
of an Urban Nuclear Power Plant." Idaho Falls, Idaho. Health Physics
Society Topical Symposium, November 1970.
Bingham, G. E., Preliminary Safety Analysis Report for the Irradiated Fuels
Storage Facility. December 1971. (ACI 113).
Bishop, W. and C. Hollister, "Seabed Disposal - Where to Look," Nuclear
Technology, 24 (December 1974).
Blomeke, J. 0., et al., "Managing Radioactive Wastes," Physics Today, 26
(August 1973) pp. 36-42.
Boch, A. L., et al., Radioactive Waste Repository Project Annual Progress
Report for Period ending Sept. 30, 1972. Oak Ridge, Tenn.: Oak Ridge
National Laboratory, December 1973). (ORNL-4824).
204
-------
Boffey, P. M., "Radiation Standards: Are the Right People Making Decisions?"
Science, 171 (Feb. 26, 1971) pp. 780-783.
Bond, W., H. Claiborne, and R. E. Leuze, "Methods for Removal of Actinides
from High-Level Wastes," Nuclear Technology, 24 (December 1974).
Bradishaw, R. L. and W. C. McClain, "Project Salt Vault: A Demonstration of
the Disposal of High-Activity, Solidified Wastes in Underground Salt
Mines," Oak Ridge, Tenn.: Oak Ridge National Laboratory, 1971. (ORNL-
4555).
Brown, D. J., "Migration Characteristics of Radionuclides Through Sediments
Underlying the Hanford Reservation," Disposal of Radioactive Wastes in
the Ground, Proceedings of a Symposium, International Atomic Energy
Agency, Vienna, 29 May - 2 June, 1967, pp. 215-228.
Brown, D. J., Status of the Ground Water Beneath Hanford Reactor Areas,
January 1962 to January 1963. April 1963. (HW 77170).
Brown, R. E. and H. G. Rupert, The Underground Disposal of Liquid Wastes at
the Hanford Works, Washington, 1950. (AEC Document HW-17088).
Chandler, J. M., The Peach Bottom Spent Fuel Element Shipping Cask Accident.
Oak Ridge, Tenn.: Oak Ridge National Laboratory, July 1972. (ORNL-TM-
3844).
Cheverton, R. D., et al., In Situ Gas Diffusion Experiments Associated with
a Radioactive Waste Repository in Bedded Salt. Oak Ridge, Tenn.: Oak
Ridge National Laboratory, January 1975. (ORNL-TM-4569).
Claiborne, H. C., Neutron-Induced Transmutation of High Level Radioactive
Waste. December 1972. (ORNL-TM-3964).
Cleary, D. P., et al., "A Preliminary Matrix Model for Quantifying and Bal-
ancing the Socio-Economic Impact of Alternative Cooling System Tech-
nologies for Nuclear Power Plants," Energy and the Environment: Cost-
Benefit Analysis, Proceedings of a short course at the Georgia Institute
of Technology, June 23-27, 1975.
Cohen, J. J., et al. , "Economic and Environmental Evaluation of Nuclear
Waste Disposal by Underground in situ Melting." Livermore, Calif.:
University of California, Lawrence Livermore Laboratory, November 1974.
(UCRL 51713).
Cohen, J. J., et al. , "Incorporation of Nuclear Waste," Nuclear Technology,
13 (April 1972).
Cohen, J. J., et al., Use of a Deep Nuclear Chimney for the In-Situ Incorpo-
ration of Nuclear Waste in Molten Silicate Rock. May 1971. (UCRL
51044).
205
-------
Cole, H. S., Disposal of Radioactive Liquid Organic Wastes at the ICPP.
Idaho Falls, Idaho: Allied Chemical Corporation, March 1972. (NTIS
Ordering No. ICP-1007).
Commander, R. E., et al., Operation of the Waste Calcining Facility with
Highly Radioactive Aqueous Waste, Report of the First Processing Cam-
paign. Idaho Falls, Idaho: U.S. Geological Survey, Water Resources
Division, June 1966. (NTIS Ordering No. IDO-14662).
Commoner, Barry. The Closing Circle. New York: Bantam, 1972.
Corley, J. P., D. M. Robertson and F. P. Brauer, "Plutonium in Surface Soil
in the Hanford Plant Environs," in Proceedings of Environmental Pluton-
ium Symposium, August 4-5, 1971. Los Alamos: Los Alamos Scientific
Laboratory, December 1971. (LA-4756).
Davis, J. J., "Accumulation of Radionuclides by Aquatic Insects," in Bio-
logical Problems in Water Pollution, Trans. 3rd Seminar (1962),
Cincinnati, Ohio: Taft Sanit. Engr. Center, 1965. (Publ. No. 999-WP-
25).
de Laguna, W., H. 0. Weeren, "Disposal of Radioactive Wastes by Hydraulic
Fracturing," Parts I, II, III, Nuclear Engineering & Design. (1966).
Deju, R. A. and W. K. Summers, Transmissivity and Hydraulic Conductivity of
Saturated Sedimentary Rocks in the Hanford Reservation. Richland,
Wash.: RAD Associates, 1975. (RAD-5).
Dkrent, D., "A General Evaluation Approach to Risk Benefit for Large Tech-
nological Systems and its Application to Nuclear Power." Los Angeles,
Calif.: University of California, School of Engineering, 1974.
Doan, P. L., "Tornadoes and Tornado Effect Considerations for Nuclear Power
Plant Structures Including the Spent Fuel Storage Pool," United
Engineers and Constructors, Inc., March 1969.
Dunster, H. J. and McLean, A. S., "The Use of Risk Estimates in Setting and
Using Basic Radiation Protection Standards," Health Physics (July 1970).
Ebbin, Steven and Raphael Kasper. Citizen Groups and the Nuclear Power
Controversy: Uses of Scientific and Technological Information. Cam-
bridge, Mass.: MIT Press, 1974.
Echo, J., "Biological Availability of Radionuclides in Calcined Waste,"
Idaho Falls, Idaho: U.S. Geological Survey, Water Resources Division,
1965. (NTIS Ordering No. 1DO-12042).
Edison Electric Institute, "Radioactive Waste Handling in the Nuclear Power
Industry," March 1960.
Feld, Bernard T., "The Menace of a Fission Power Economy," Science and Public
Affairs, Bulletin of the Atomic Scientists, (April 1974) pp. 32-34.
206
-------
Fielding, J. R., Safety Review Document for Nuclear Reactor Testing Station
Radioactive Waste Burial Ground, ANCR.
Foster, R. F. and J. K. Soldat, "Evaluation of the Exposure Resulting from
the Disposal of Radioactive Wastes into the Columbia River," Disposal
of Radioactive Wastes into Seas, Oceans and Surface Waters. 1966
Symposium, International Atomic Energy Agency, Vienna, 1966, pp. 683-
695.
Fraas, A. P. and Steiner, D., "Preliminary Observations on the Radiological
Implications of Fusion Power," Nuclear Safety, 1_3 (Sept. - Oct. 1972)
pp. 353-362.
Freke, A. M., "A Model for the Approximate Calculation of Safe Rates of Dis-
charge of Radioactive Wastes into Marine Environments," Health Physics,
13 (1965) p. 43.
Cast, P. F. (Argonne National Laboratory), "Divergent Public Attitudes
Toward Nuclear and Hydroelectric Plant Safety." Presented at American
Nuclear Society Meeting, Chicago, 111., June 1973.
Geesaman, D. P., "Plutonium and Public Health," Livermore, Calif.: Lawrence
Livermore Laboratory, April 1970. (GT-121-70).
Gillette, Robert, "Radiation Spill at Hanford: The Anatomy of an Accident,"
Science, 181 (August 24, 1973) pp. 728-730.
Gillette, Robert, "Radiation Standards: The Last Word or at Least a Defini-
tive One," Science, 178 (Dec. 1, 1972) pp. 966-967, 1012.
Glauberman, H. , W. R. Boatman and A. J. Breslin, "Studies of the Signifi-
cance of Surface Contamination," in B. R. Fish, Ed., Surface Contamina-
tion, New York: Pergamon Press, 1967.
Glenn, J. L., Relations Among Radionuclide Content and Physical,Chemical and
Mineral Characteristics of Columbia River Sediments, Portland, Oregon:
U.S. Geological Survey, Water Resource Division, 1971.
Gruenstein, P. and R. H. Sandier, "Power from Fission: Potential for
Catastrophe," The Progressive, 37 (November 1973) pp. 36-41.
Hammond, J. D., et al., Safety Analysis Report for Peach Bottom I Core 1
Fuel Storage Facility. Idaho Falls, Idaho: Idaho Nuclear Corporation,
June 1971. (NTIS Ordering No. IN-1465).
Harvey, R. W. and W. Schmidt, "Radioactive Waste Management at Hanford,"
First Revision, Atlantic Richfield Hanford Company, March 1971.
Hill, M. N., Ed., The Sea. New York: Wiley, 1974.
207
-------
Hirshleifer, J., T. Bergstrom and E. Rappaport, "Applying Cost-Benefit Con-
cepts to Projects Which Alter Human Mortality — A General Evaluation
Approach to Risk-Benefit for Large Technological Systems and Its Appli-
cation to Nuclear Power." Los Angeles, Calif.: University of
California, School of Engineering and Applied Sciences (UCLA-ENG-7478),
November 1974.
Hogg, G. W., et al., A Survey of NRTS Waste Management Practices, Volumes I
and II, September 1971. (ACI-104).
Holden, C., "Nuclear Waste: Kansans Riled by AEC Plans for Atom Dump,"
Science, 172 (April 16, 1972) p. 249.
Howard, R. A., "Social Decision Analysis," Proceedings of the Institute of
Electrical and Electronics Engineers, 63 (March 1975).
Hubbell, D. W. and J. L. Glenn, "Distribution of Radionuclides in Bottom
Sediments of the Columbia River," Open File Report, U.S. Geological
Survey, 1971.
Hyland, R. E., et al., Feasibility Study of Space Disposal of Nuclear Waste,
NASA-AEC Study Group, NASA TM-X-2912, vol. 2.
Hyland, R. E., et al. , Study of Extraterrestrial Disposal of Radioactive
Wastes, Part II, "Preliminary Feasibility Screening Study of Extra-
terrestrial Disposal of Radioactive Wastes in Concentrations, Matrix
Materials and Containers Designed for Storage on Earth," Cleveland,
Ohio: Lewis Research Center. October 1972. (NASA-TM-X-68147).
Imbert, J. C., and F. Pacaud, "Study of Diffusion/Leaching Phenomena with
Glasses, Application of Risk Evaluation to Long-Term Storage," CEA
Centre de Marcoule, 30- Bagnols-sur-Ceze (France), August 1974.
International Commission on Radiological Protection. Recommendations.
(Adopted September 17, 1966) New York: Pergamon Press, 1966.
Joskow, P. L., "Approving Nuclear Power Plants: Scientific Decisionmaking
or Administrative Charade?" Bell Journal of Economics & Management
Science, 5 (Spring 1974).
Kneese, A., "The Faustian Bargain," in Resources, No. 44, (Washington,
D.C.: Resources for the Future, September 1973).
Krumholz, L. A. and R. F. Foster, "Accumulation and Retention of Radio-
activity from Fission Products and Other Radioactive Materials by
Freshwater Organisms," in The Effects of Atomic Radiation on Oceano-
graphy and Fisheries. 1957. pp. 88-95 (National Academy of Sciences -
National Research Council, Publication No. 551).
Kubo, A. S., Technology Assessment of High-Level Nuclear Waste Management.
Sc.D. Thesis, Department of Nuclear Engineering, Massachusetts Insti-
tute of Technology, April 1973.
208
-------
Kubo, A. S., and D. J. Rose, "On Disposal of Nuclear Waste," Science, 182
(December 21, 1973) pp. 1205-1211.
Langham, W. H., The Problem of Large Area Plutonium Contamination. U.S.
Department of Health, Education and Welfare, Public Health Services,
Seminar Paper No. 002, December 6, 1968.
Larson, D. E., High-Level Liquid Radioactive Waste Management Program. U.S.
Atomic Energy Commission Report, April 1972. (ARH 2400).
Lash, Terry R. "Prepared Testimony Before The Energy Facility Siting
Council of Oregon," January 11, 1977. (mimeo).
Lave, L. B. and L. C. Freeburg, "Health Effects of Electricity Generation
from Coal, Oil and Nuclear Fuel," Nuclear.Safety, (Sent. - Oct. 1973).
Lawless, Edward. Technology and Social Shock. Final report for the National
Science Foundation. New Brunswick, N.J.: Rutgers University Press,
1977.
Lennemann, W. L., "Management of Radioactive Aqueous Wastes from U.S. Atomic
Energy Commission's Fuel Reprocessing Operations, Experience and Plan-
ning," in Symposium on the Management of Radioactive Wastes, Paris,
France, November 27 - December 1, 1972.
Lewis, R. S., The Nuclear Power Rebellion. New York: The Viking Press,
1972.
Lewis, R. S., "The Radioactive Salt Mine," Science and Public Affairs,
Bulletin of the Atomic Scientists, (June 1971) p. 27.
Lohse, G. E., Safety Analysis Report for the ICPP High Level Solid Radio-
active Waste Storage Facilities. Idaho Falls, Idaho: Allied Chemical
Corporation, January 1972. (NTIS Ordering No. ICP-1005).
Lohse, G. E., Second Processing Campaign in the Waste Calcining Facility.
Idaho Falls, Idaho: Idaho Nuclear Corporation, March 1970. (NTIS
Ordering No. IN-1344).
Louis Harris and Associates, Inc. A Survey of Public and Leadership Atti-
tudes Toward Nuclear Power Development in the United States. Conducted
for Ebasco Services, Inc., New York, August 1975.
Marcus, M., "The Economic Benefits of Nuclear Power Plants," Public Utilities
Fortnightly, (June 20, 1974) pp. 26-30.
Markham, 0. D., "Environmental and Radiological Monitoring at the National
Reactor Testing Station During FY 1973," Radiation Data and Reports,
15, (May 1974), (U.S. Environmental Protection Agency Publication).
209
-------
Maxfield, H. L., Radioactive Contamination in Unplanned Release to Ground
within the Chemical Separation Area Control Zone through 1972. Part 4.
Richland, Wash.: Atlantic Richfield Hanford Corporation, April 3, 1973.
(NTIS Ordering No. ARH-2757).
McCarthy, J. F., Jr., et al., "Concepts for Space Disposal of Nuclear Waste."
Cambridge, Mass.: Massachusetts Institute of Technology, October 1972.
(Graduate Student Summary).
McClain, W. C., "Disposal by Hydraulic Fracturing." Oak Ridge, Tenn.: Oak
Ridge National Laboratory, 1967. (NTIS Ordering No. ORNL-TM-1846).
McClain, W. and A. Bock, "Disposal of Radioactive Waste in Bedded Salt Forma-
tions," Nuclear Technology, 24 (December 1974).
McGrath, P. E., "Radioactive Waste Management Potentials & Hazards from a
Risk Point of View," EURFNR-1204, Kernforschungszentrum Karlsruhe (F. R.
Germany), June 1974.
Mellor, M., "Ice Flow in Antarctica," Journal of Glaciology, 3 (1959).
Miklin, Philip P. "Environmental Hazards of Nuclear Wastes," Science and
Public Affairs, The Bulletin of the Atomic Scientists, 30 (April 1974).
Mitre Corporation. Proceedings of Quantitative Environmental Comparison of
Coal and Nuclear Electric Generators and their Associated Cycles Work-
shop, Washington, D.C.: Mitre Corporation, 1975. Vol. 1 (MTR-7010).
Mommsen, J., "A Probabilistic Uranium Forecast," Nuclear Exchange Corp.,
Paper Presented at Atomic Industrial Forum, March 1975.
Murphy, Arthur W. "The National Environmental Policy Act and The Licensing
Process: Environmental List Magna Carta or Agency Coup de Grace?,"
Columbia Law Review, 72 (October 1972) pp. 962-1007.
National Academy of Sciences. Disposal of Low-Level Radioactive Waste into
Pacific Coastal Waters, 1962 (Publication 985).
National Academy of Sciences- National Research Council. "The Disposal of
Radioactive Waste on Land," Report of the Committee on Waste Disposal of
the Division of Earth Sciences, September 1957 (NAS-NRC Publication 519).
National Academy of Sciences- National Research Council. Committee on
Geological Aspects of Radioactive Waste Disposal, Report to Division
of Reactor Development and Technology, USAEC, May 1966.
National Academy of Sciences, Division of Earth Sciences. "Report of
Committee on Geologic Aspects of Radioactive Waste Disposal, May 1966.
210
-------
National Academy of Sciences, Division of Medical Sciences. "Effects on
Populations of Exposure to Low Levels of Ionizing Radiation." Report
of Advisory Committee on Biological Effects of Ionizing Radiations.
Washington, D.C.: National Academy of Sciences, November 1972.
National Broadcasting Company. "Danger: Radioactive Waste," Transcript of
a Broadcast, January 26, 1977.
Natural Resources Defense Council. Alternative Technology Options; Commen-
tary on WASH-1535, by A. R. Tamplin, et al. Palo Alto, Calif.: Natural
Resources Defense Council.
Natural Resources Defense Council. Citizens' Guide: The National Debate on
the Handling of Radioactive Wastes from Nuclear Power Plants, by Terry
R. Lash, John E. Bryson, and Richard Cotton. Palo Alto, Calif.:
Natural Resources Defense Council, 1975.
Natural Resources Defense Council. Transportation of Radioactive Materials.
Comments on WASH-1535. Palo Alto, Calif.: Natural Resources Defense
Council.
Nelson, I. C. and R. F. Foster. "Ringold Farms — A Hanford Environmental
Study," Health Physics, 11 (1965) pp. 397-401.
Nichols, J. P., et al., Projections of Fuel Reprocessing Requirements and
High-Level Solidified Wastes from the U.S. Nuclear Power Industry.
Oak Ridge, Tenn.: Oak Ridge National Laboratory (ORNL-TM-3965).
Novick, S., The Careless Atom. Boston: Houghton Mifflin, 1969.
Nye, J. F., "The Response of Glaciers and Ice Sheets to Seasonal and Cli-
matic Changes," Proceedings of the Royal Society, 256 (1960).
O'Keefe, D. R., Feasibility of Using an Orbiting Accelerator to Eject Radio-
active Waste Products into Space, Gulf-RT-Cl2457, Interim Summary
Report, July 1973.
Osloond, J. H., Radioactive Waste Disposal Data for the National Reactor
Testing Station, Idaho. Idaho Falls, Idaho: U.S. Geological Survey,
Water Resources Division, Supplement No. 2, November 1967. (NTIS
Ordering No. IDO-12040).
Oswald, G. K. A., and G. de Q. Robin, "Lakes Beneath the Antarctic Ice
Sheet," Nature, 245 (October 1973).
Perona, J. J., et al., Design and Safety Considerations of Shipping Solidi-
fied High-Level Radioactive Wastes. Oak Ridge, Tenn.: Oake Ridge
National Laboratory, December 1970. (ORNL-TM-2971).
Pittman, F. K., Plan for the Management of AEC-Generated Radioactive Wastes,
January 1972. (WASH-1202).
211
-------
Pruter, A. T. and D. L. Alverson, Eds., The Columbia River Estuary and
Adjacent Ocean Waters - Bioenvironmental Studies. Seattle: University
of Washington Press, 1972.
Pugwash Newsletter, 11, Nos. 1 & 2 (July and October, 1973) p. 23.
Reynolds, W. C. The California Nuclear Initiative: Analysis and Discussion
of Issues. Stanford, Calif.: Stanford University Institute for Energy
Studies, April 1976.
Rhodes, Richard. "Delusions of Power: The Benefits, Costs and Risks of
Nuclear Energy," The Atlantic, 237 (June 1976) pp. 35-48.
Robertson, D. E., et al., "Transport and Depletion of Radionuclides in the
Columbia River," Vienna, International Atomic Energy Agency, 1973.
Robertson, J. B., et al., The Influence of Liquid Waste Disposal on Geo-
chemistry of Water at the National Reactor Testing Station, Idaho,
1952-1970. Idaho Falls, Idaho: U.S. Geological Survey, Water Resources
Division, 1974. (NTIS Ordering No. IDO-22053).
Ronca, L. B., "Minimum Length of Time of Frigid Conditions in Antarctica as
Determined by Thermoluminescence," American Journal of Science, 262
(June 1964).
Sagan, L. A., "Human Costs of Nuclear Power," Science, 177 (August 11,
1972) pp. 487-493.
Scheinman, L., "Safeguarding Nuclear Materials," Science and Public Affairs,
Bulletin of the Atomic Scientists. (April 1974) pp. 34-36.
Schmalz, B. L. and W. L. Polzer, Radionuclide Distribution in the Regolith
as a Result of Waste Disposal at the National Reactor Testing Station.
Idaho Falls, Idaho: U.S. Geological Survey, Water Resources Division,
June 1970 (NTIS Ordering No. IDO-12072).
Schneider, K. J., "Radioactive Waste from Nuclear Power in the U.S.A.,"
AIME Environmental Quality Conference, Washington, D.C., June 7-9, 1971.
Schneider, K. J., "Solidification and Disposal of High-Level Radioactive
Wastes in the United States, A Critical Review," Reactor Technology, 13
(Winter 1970-71).
Shappert, L. B., W. A. Brobst, J. W. Langhaar, and J. A. Sisler, "Probability
and Consequences of Transportation Accidents Involving Radioactive
Material Shipments in the Nuclear Fuel Cycle," Nuclear Safety, 14 (Nov.-
Dec., 1973).
Spangler, Miller B., "The Basis for Standards in Cost-Benefit Analysis in
Nuclear Power Plant Licensing Actions," paper presented in: Environ-
mental Impact Statements for Nuclear Power Plants. Atlanta, Ga.:
Georgia Institute of Technology, November 29, 1973.
212
-------
Spangler, Miller B., "The Treatment of Risk and Uncertainty in Cost-Benefit
Analysis of Environmental Impact Statements for Nuclear Power Plant
Licensing Actions," Washington, D.C.: USAEC, 1974.
Starr, Chauacey, "Benefit-Cost Relationships to Socio-Technical Systems,"
in: Environmental Aspects of Nuclear Power Stations, Vienna, Interna-
tional Atomic Energy Agency, 1971.
Starr, Chauncey, "Social Benefit vs. Technological Risk - What is our Society
Willing to Pay for Safety?" Science, 165 (1969).
"A Study of Social Costs for Alternative Means of Electrical Power Generation
for 1980 and 1990," Summary Report. February 1973. (ANL 8092).
Sverdrup, H. U., M. W. Johnson and R. H. Fleming, The Oceans. New York:
Prentice-Hall, 1952.
Terrill, J. G. "Cost-Benefit Estimates for the Major Sources of Radiation
Exposure," American Journal of Public Health, 62 (July 1972) pp. 1008-
1013.
Tversky, A., and D. Kahneman, "Judgement Under Uncertainty: Heuristics and
Biases," Science, 185 (September 27, 1974).
Union of Concerned Scientists. The Nuclear Fuel Cycle; A Survey of the
Public Health, Environmental and National Security Effects of Nuclear
Power. Revised edition. Cambridge, Mass.: MIT Press, 1975.
United Nations Association of the U.S.A., Safeguarding the Atom: A Soviet
American Exchange. New York, N.Y.: U.N. Association of the U.S.A.,
July 1972.
U.S. Atomic Energy Commission. Considerations in the Long-Term Management
of High-Level Radioactive Wastes, Oak Ridge, Tenn.: Oak Ridge National
Laboratory, 1972. (ORNL 4762).
U.S. Atomic Energy Commission. "Contaminated Soil Removal Facility, EIS,"
April 1972.
U.S. Atomic Energy Commission. Draft Environmental Statement, A Waste
Calcination Facility, National Reactor Testing Station, Idaho,
December 1972 (WASH-1531).
U.S. Atomic Energy Commission. Environmental Statement, Rover Fuels Process-
ing Facility, National Reactor Testing Station, Idaho, April 1972 (WASH-
1512).
U.S. Atomic Energy Commission. Environmental Statement, Waste Calcine
Storage Bins, National Reactor Testing Station, Idaho, April 1973
(WASH-1529).
213
-------
U.S. Atomic Energy Commission. Environmental Survey of Transportation of
Radioactive Materials to and from Nuclear Power Plants, Washington, D.C.
Government Printing Office, 1972. (WASH-1238).
U.S. Atomic Energy Commission. Hanford Radioactive Waste Management Plans,
PWM-530, Rev. 1, 1973.
U.S. Atomic Energy Commission. Hanford Waste Management Operations Draft
Environmental Impact Statement, ERDA Hanford Operations, September 1974.
(WASH-1538).
U.S. Atomic Energy Commission. High-Level Radioactive Waste Management
Alternatives, May 1974. (WASH-1297).
U.S. Atomic Energy Commission. Management of Commercial High-Level and
Trans-Uranium-Contaminated Radioactive Wastes, Environmental Statement,
September 1974. (WASH-1539).
U.S. Atomic Energy Commission. Plans of the Management of ABC-Generated
Radioactive Waste, July 1973. (WASH-1202).
U.S. Atomic Energy Commission. Potential Containment Failure Mechanisms and
Their Consequences at a Radioactive Waste Repository in Bedded Salt in
New Mexico, Oak Ridge, Term.: Oak Ridge National Laboratory, October
1974. (ORNL-TM-4639).
U.S. Atomic Energy Commission. Radioactive Waste Repository, Lyons, Kansas,
June 1971. (WASH-1503).
U.S. Atomic Energy Commission. Reactor Fuel Cycle Costs for Nuclear Power
Evaluation, Oak Ridge, Term.: Oak Ridge National Laboratory, 1971.
(WASH-1099).
U.S. Atomic Energy Commission. Rules and Regulations, Title 10 - Atomic
Energy, Part 2, Rules of Practice, October 27, 1966.
U.S. Atomic Energy Commission. Waste Management Operations, Hanford Reser-
vation, Richland, Washington: Environmental Statement, Vol. I, Vol. II
and Appendices. Washington, D.C., September 1974. (WASH-1538).
U.S. Code of Federal Regulations. Title 40, Part 1500.13, 38. "Guidelines,
Council on Environmental Quality," Federal Register (147) (August 1,
1973).
U.S. Code of Federal Regulations. Title 49, Parts 170-189, and Title 14,
Part 103, "Hazardous Materials Regulations."
U.S. Code of Federal Regulations. Title 46, Part 146, "Transportation or
Storage of Explosives or Other Dangerous Articles or Substances, and
Combustible Liquids Onboard Vessels."
214
-------
U.S. Comptroller General. Observations Concerning the Management of High-
Level Radioactive Waste Material, Report to the Joint Committee on
Atomic Energy, May 1968.
U.S. Comptroller General. .Progress and Problems in Programs for Managing
High-Level Radioactive Wastes, Report to the Joint Committee on Atomic
Energy, January 29, 1971.
U.S. Congress. House, Committee on Interior and Insular Affairs. Nuclear
Waste Disposal in Michigan, Oversight Hearing before the Subcommittee
on Energy and the Environment. 94th Congress, 2nd Session, 1976.
U.S. Congress, Senate. Clean Air Amendment of 1970, Public Law 91-604. 84
Stat. 1676. U.S. Code, Congressional and Administrative News. 91st
Congress, 2nd Session, 1970. Vol. 1. Washington, D.C.: Government
Printing Office, 1971.
U.S. Department of Commerce, "Radioactive Waste Disposal in the Ocean,"
National Bureau of Standards Handbook 58, August 25, 1954.
U.S. Department of the Interior. Federal Water Pollution Control Adminis-
tration. Northwest Region. Report on the Waste Treatment and Disposal
Operations at the 100 and 300 Areas. Richland, Wash.: November 1969.
(Hanford Project).
U.S. Energy Research and Development Administration. Alternatives for
Managing Wastes from Reactors and Post-Fission Operations in the LWR
Fuel Cycle. Report coordinated by Battelle, Pacific Northwest Labora-
tories for the Division of Nuclear Fuel Cycle and Production, U.S.
Energy and Research Administration. 1976 (ERDA 76-43).
U.S. Environmental Protection Agency. Office of Radiation Programs.
Division of Criteria & Standards. "Draft Report of the Special Studies
Group: Estimates of Ionizing Radiation Doses in the United States,
1960-2000," Rockville, Maryland, June 1971.
U.S. Federal Radiation Council. "Background Material for the Development
of Radiation Protection Standards, Staff Reports 1, 2, 5, 7: Protec-
tive Action Guides for Strontium-89, Strontium-90, and Cesium-137."
Washington, D.C.: Government Printing Office, 1960, 1961, 1964, 1965.
U.S. Federal Radiation Council. Federal Radiation Guide, Reports Nos. 1-7.
U.S. General Accounting Office. Isolating High-Level Radioactive Waste from
the Environment: Achievements, Problems, and Uncertainties. December
1974. (RED-75-309).
U.S. General Accounting Office. Observations Concerning the Management of
High-Level Radioactive Waste Material. May 29, 1968. (B-164052).
U.S. General Accounting Office. Progress and Problems in Programs for Man-
aging High-Level Radioactive Wastes. January 29, 1971. (B-164052).
215
-------
U.S. Nuclear Regulatory Commission. Environmental Survey of the Reprocess-
ing and Waste Management Portions of the LRW Fuel Cycle, October 1976.
(NUREG 0116: Supplement to WASH-1248).
Walker, E., "Subsurface Geology of National Reactor Testing Station Idaho,"
Geologic Survey Bulletin 113-E (1964).
Weber, R., "Buried for Geological Periods of Time: Deep Disposal of Liquid
Wastes," Bundesanstalt fur Bodenforschung, Hanover - Buchholz (F. R.
Germany), Umwelt, 4 (August 1974) pp. 42-43.
Weertman, J., "Effect of a Basal Water Layer on the Dimensions of Ice
Sheets," Journal of Glaciology (June 1966) p. 6.
Weinberg, Alvin M., "Social Institutions and Nuclear Energy," Science, 177
(7 July 1972) pp. 27-35.
Weinberg, Alvin M., "The Moral Imperatives of Nuclear Energy," Nuclear News
(December 1971) pp. 33-37.
White, Gilbert F. and J. Eugene Haas. Assessment of Research on Natural
Hazards. Cambridge, Mass.: MIT Press, 1975.
Wieland, J. A., et al., The Fifth Processing Campaign in the Waste Calcining
Facility FY 1972. Idaho Falls, Idaho: Allied Chemical Corporation,
June 1973. (NTIS Ordering No. ICP-1021).
Willrich, M., Global Politics of Nuclear Energy. New York: Praeger Publish-
ers, 1971.
Willrich, M., Ed., International Safeguards and the Nuclear Industry. Balti-
more: Johns Hopkins University Press, 1973.
Willrich, M., and T. B. Taylor, Nuclear Theft; Risks and Safeguards (A
Report to the Ford Foundation Energy Policy Study), Cambridge, Mass.:
Ballinger Publishing Company, 1974.
Wilson, D. 0. and J. F. Cline, "Removal of Plutonium-239, Tungsten-185 and
Lead-210 from Soils," Nature, 209 (1966) pp. 941-942.
Wolfe, B. and R. W. Lambert, "The Back End of the Fuel Cycle," General
Electric, 1975.
Yadigaroglu, G., A. G. Reinking, and V. E. Schrock, "Spent Fuel Transporta-
tion Risks," Nuclear News (November 1972) pp. 71-75.
Zellner, E. J., D. F. Sanders, and E. E. Angino, "Putting Radioactive Wastes
on Ice," Science and Public Affairs, Bulletin of the Atomic Scientists
(January 1973).
216
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
i. REP.OJRT.ua.
EPA-600/5-78-021
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Social Decision Making for High Consequence, Low
Probability Occurrences.
5. REPORT DATE
October 1978
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO,
9. PERFORMING ORGANIZATION NAME AND ADDRESS
National Planing Association, Washington, D. C. and
Haldi Associates, Inc., New York, N.Y.
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
EPA Contract No. 68-01-3228
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Research and Development
EPA, Corvallis Environmental Research Laboratory
Corvallis, Oregon 97330
13. TYPE OF REPORT AND PERIOD COVERED
Final 1975-1977
14. SPONSORING AGENCY CODE
EPA/600/02
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This study deals with the process of reaching social decisions that involve low-
probability, high conseguence outcomes. It is in 3 major parts. Part I reviews
the two main classes of criteria proposed for social decisions: (1) market mechanism
and cost-benefit analysis and (2) the approaches of Rawls and Buchanan to arrive at a
social consensus. The authors propose an eight element criteria for evaluating a
social decision process capable of application in judicial, legislative, academic,
and managerial situations.
Parts II and III are case studies of different decision procedures. The former
inspects the administrative law procedure using the U.S. EPA chlordane/heptachlor
suspension hearings for example. The legal procedure is concluded to be a forward-
looking decision process in an effort to increase social acceptability of the out-
come. The latter examines the disjointed decision procedure concerning the disposal
of nuclear waste. In both cases the eight criteria proposed in Part I nevertheless
provide a useful tool for evaluating the process.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI F-'icld/Group
Economic Impact Analysis
Environmental Economics
Benefit-Cost Analysis
Decision Making
Environmental Law
Environmental Policy
Decision Theory
Risk Analysis
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
230
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION i s OBSOLETE
US GOVERNMENT PRINTING OFFICE: 1978—798/084-2-1
217
-------
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Environmental Research Information Center
Cincinnati, Ohio 45268
OFFICIAL BUSINESS
PENALTY FOR PRIVATE USE S3OO
AN EQUAL OPPORTUNITY EMPLOYER
POSTAGE AND FEES PAID
U S ENVIRONMENTAL PROTECTION AGENCY
EPA-335
Special Fourth-C lass R ate
Book
If your address is incorrect, please change on the above label
tear off; and return to the above address.
If you do not desire to continue receiving these technical
reports, CHECK HERE L~3, tear off label, and return it to the
above address.
------- |