EPA-2 3 0-12-85-019
September 1985
METHODS DEVELOPMENT FOR ENVIRONMENTAL
CONTROL BENEFITS ASSESSMENT
Volume I
MEASURING THE BENEFITS OF CLEAN AIR AND WATER
by
Allen V. Kneese
Resources for the Future, Inc.
Washington, D.C. 20036
Based on Research Reports Whose Principal Authors Are:
Richard Adams - University of Wyoming
Ralph d'Arge - University of Wyoming
Shaul Ben-David - University of New Mexico
David Boldt - SRI International
David Brookshire - University of Wyoming
Richard Carson - Resources for the Future
Ronald Cummings - University of New Mexico
Thomas Crocker - University of Wyoming
Maureen Cropper - University of Maryland
Shelby Gerking - University of Wyoming
Leonard Gianessi - Resources for the Future
Michael Hazilla - Resources for the Future
Raymond Kopp - Resources for the Future
Edna Loehman - SRI International
Robert Mitchell - Resources for the Future
John Mullahy - Resources for the Future
Henry Peskin - Resources for the Future
Paul Portney - Resources for the Future
Clifford Russell - Resources for the Future
William Schulze - University of Wyoming
Mark Sharefkin - Resources for the Future
Mark Thayer - San Diego State University
William Vaughan - Resources for the Future
USEPA Grants No. R810466-01-0
R805059- 01-0
Project officer
Dr. Alan Carlin
Office of Policy, Planning and Evaluation
U.S. Environmental Protection Agency
Washington, D.C. 20460
OFFICE OF POLICY ANALYSIS
OFFICE OF POLICY, PLANNING AND EVALUATION
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
-------
OTHER VOLUMES IN THIS SERIES
Volume 2, Six Studies of Health Benefits from Air Pollution Control ,
EPA-2 3 012-85-020.
This volume contains six statistical epidemiology studies. They
show that large associations between health and current levels of air
pollution are not robust with respect to the statistical model
specification either for mortality or morbidity. They also find that
significant relationships, mostly small, occasionally appear.
Volume 3, Five Studies on Non-Market Valuation Techniques, EPA-230-12-
85-021.
This volume presents analytical and empirical comparisons of
alternative techniques for the valuation of non-market goods. The
methodological base of the survey approach - directly asking individuals
to reveal their preference in a structured hypothetical market - is
examined for bias, replication, and validation characteristics.
Volume 4, Measuring the Benefits of Air Quality Changes in the San
Francisco Bay Area: Property Value and Contingent Valuation Studies,
EPA-2 3 0-12-85-022.
This volume replicates a property value study conducted in the Los
Angeles Basin for the San Francisco Bay area. A taxonomy series of air
quality types and socioeconomic typologies are defined for cities in the
area to examine how property values vary with pollution levels. The
contingent valuation method surveys individuals, directly asking their
willingness to pay for changes in air quality. The survey method yields
benefit values that are about half the property value benefits in both
the Bay area and Los Angeles.
Volume 5, Measuring Household Soiling Damages from Suspended
Particulates: A Methodological Inquiry, EPA 230-12-85-023.
This volume estimates the benefits of reducing particulate matter
levels by examining the reduced costs of household cleaning. The
analysis considers the reduced frequency of cleaning for households that
clean themselves or hire a cleaning service. These estimates were
compared with willingness to pay estimates for total elimination of air
pollutants in several U.S. cities. The report concludes that the
willingness-to-pay approach to estimate particulate-related household
soiling damages is not feasible.
Volume 6, The Value of Air Pollution Damages to Agricultural Activities
in Southern California, EPA-230-12-85-024.
This volume contains three papers that address the economic
implications of air pollution-induced output, input pricing, cropping,
and location pattern adjustments for Southern California agriculture.
The first paper estimates the economic losses to fourteen highly valued
vegetable and field crops due to pollution. The second estimates
earnings losses to field workers exposed to oxidants. The last uses an
econometric model to measure the reduction of economic surpluses in
Southern California due to oxidants.
-------
Volume 7, Methods Development for Assessing Acid Deposition Control
Benefits, EPA-230-12-85-025.
This volume suggests types of natural science research that would
be most useful to the economist faced with the task of assessing the
economic benefits of controlling acid precipitation. Part of the report
is devoted to development of a resource allocation process framework for
explaining the behavior of ecosystems that can be integrated into a
benefit/cost analysis, addressing diversity and stability.
Volume 8, The Benefits of Preserving Visibility in the National
Parklands of the Southwest, EPA-230-12-85-026.
This volume examines the willingness-to-pay responses of individuals
surveyed in several U.S. cities for visibility improvements or
preservation in several National Parks. The respondents were asked to
state their willingness to pay in the form of higher utility bills to
prevent visibility deterioration. The sampled responses were
extrapolated to the entire U.S. to estimate the national benefits of
visibility preservation.
Volume 9, Evaluation of Decision Models for Environmental Management,
EPA-2 3 012-85-0@7.
This volume discusses how EPA can use decision models to achieve the
proper role of the government in a market economy. The report
recommends three models useful for environmental management with a focus
on those that allow for a consideration of all tradeoffs.
Volume 10, Executive Summary, EPA-230-12-85-028.
This volume summarizes the methodological and empirical findings
of the series. The consensus of the empirical reports is the benefits
of air pollution control appear to be sufficient to warrant current
ambient air quality standards. The report indicates the greatest
proportion of benefits from control resides, not in health benefits, but
in aesthetic improvements, maintenance of the ecosystem for recreation,
and the reduction of damages to artifacts and materials.
II
-------
FOREWORD
This volume is one of the reports prepared by research
institutions under cooperative agreements with the Economic Research
Program of the United States Environmental Protection Agency (EPA). The
purpose of the Program is to carry out economic research that will
assist EPA in carrying out its mission. Until very recently, most
research sponsored by the Program @t to improve the methods and data
available for determining the economic benefits of pollution control,
thereby assisting EPA and other Federal Agencies responsible for
preparing benefit-cost analyses of programs and regulations. Such
benefit-cost analyses are required as part of the Regulatory Impact
Analyses mandated for most major Federal regulations by Executive Order
12291. The availability of improved methods and data will make it
possible for EPA and other Agencies to determine more accurately the
economic efficiency of their regulations and programs. Very recently,
the scope of the Program has been expanded to include a broader range of
research on increasing the economic efficiency of pollution control.
The Economic Research Program was a part of the office of Research
and Development (ORD) until early 1983, when it was transferred to what
is now the Office of Policy, Planning and Evaluation. The cooperative
agreements under which this volume was prepared were concluded while the
Program was still in ORD; accordingly, ORD's important contribution
should be recognized.
This volume is one of a series under the title Methods Development
for Environmental Control Benefits Assessment prepared mainly under
cooperative agreement R805059 with the University of Wyoming, although
several of the individual volumes were completed under later cooperative
agreements or under subagreements with other institutions. Each of the
other volumes in the series is listed on the front and back inside
covers of this volume. The overall purpose of the series is to report
significant research results achieved under the cooperative agreement.
The purpose of the agreement was to develop improved methods for
assessing environmental benefits, with emphasis on air pollution
benefits. An earlier series of interim reports prepared under the same
cooperative agreement was published by EPA in 197 9 under the series
title of Methods Development for Assessing Air Pollution Control
Benefits with report numbers EPA-600/5-79-001a through OOle.
This volume is a nontechnical summary of most of the economic
benefits research funded by EPA at the Universities of Wyoming and New
Mexico and at Resources for the Future over the period 1976-83. As
such, it represents an overview of much of the research funded under the
Economic Research during this period, as well as of a few studies funded
by the Office of Air Quality Planning and Standards. Although
originally prepared under cooperative agreement R805059, this summary
was extended under R810466 with Resources for the Future. A number of
the research studies summarized in this volume (and listed in the
Bibliography) are part of this same series.
Alan Carlin
Office of Policy, Planning
and Evaluation (PM-220)
Washington, D.C.
Ill
-------
ABSTRACT
This volume is a nontechnical discussion of the work of a number of
scholars located at Resources for the Future, the University of Wyoming,
the University of New Mexico, and the University of Chicago. The focus
of these efforts was to develop improved methods for the economic
evaluation of environmental improvements or maintenance. The work was
sponsored by the U.S. Environmental Protection Agency by a sustained
program in this area of research. The studies centered on two broad
approaches. The first involves methods based on actual behavior with
respect to environmental goods. These include travel to recreational
opportunities of varying quality, prices paid for houses in environments
of different quality, decisions about farm crops depending upon how they
are affected by air pollution. While a certain confidence adheres to
methods based on actual decisions because of their nonhypothetical
nature, these methods are not applicable to all environmental benefits.
For example, they are not suitable for evaluating visibility effects of
air pollution in large landscapes or to a category of benefits termed
nonuser or intrinsic. The latter are benefits to people who have a
preference for environmental quality in situations in which they do not
actually participate. For example, people may value good water quality
for the nation as a whole even though they do not recreate in natural
waters. Accordingly, resort is made to a set of methods called
contingent valuation. These methods rely on questioning respondents
about their willingness to pay for various hypothetical changes in
environmental quality. While doubts about the accuracy of these methods
necessarily arise because of their hypothetical nature, the research
reported in this volume suggests that the identified sources of possible
bias can be controlled for by careful questionnaire design. There
remain, however, some questions for future research. In particular, the
matter of how to get respondents to evaluate their replies in terms of
their overall budgetary situations invites further inquiry. While the
central focus of the research reported in this volume was methods
development, some broad insights concerning the quantitative benefits
from environmental maintenance or improvement also emerged.
IV
-------
CONTENTS
FOREWORD
ABSTRACT
PART I BASICS 1
Chapter 1 Introduction to Benefit-Cost Analysis 2
Chapter 2 What are Economic Benefits? 7
Introduction 7
Individual Demand 7
Aggregate Demand 10
Private Goods and Public Goods 11
Compensation 13
Chapter 3 Links Between Actions That Affect the Environment
and Effects on Humans 14
Introduction 14
Health Linkages 14
Visual Quality 15
Links to Agricultural Productivity 15
Links in Watercourses 16
Aquatic Ecosystem Linkages 17
Materials Damage 18
Groundwater Linkages 18
Chapter 4 Problems of Assigning Economic Values 2 0
Introduction 2 0
Valuing Risk to Health 2 0
Morbidity 22
Visual Perception 23
Introduction 23
Strategic Bias 24
Information Bias 24
Starting Point Bias 24
Hypothetical Bias 25
Conclusions About Bidding Game Bias 25
Water-based Recreation 26
Valuing Agricultural Impacts 28
Residential Property Values--A Summary Measure?.. 30
PART II CASE STUDIES 32
PART IIA URBAN AIR POLLUTION 3 3
Chapter 5 Aggregate Epidemiology--The Sixty Cities Study... 34
The Sixty Cities Study 34
Chapter 6 Disaggregate Epidemiology and Morbidity 41
V
-------
Chapter 7 Air Quality Benefits in the South Coast Air Basin
and in San Francisco
The Basic Studies
An Illustrative Benefit-Cost Analysis
Chapter 8 Air Quality, Wages, and National Benefits from
Urban Air Pollution Control
PART I IB RURAL AND REGIONAL AIR AND WATER POLLUTION
Chapter 9 Air Quality Benefits to Southern California
Agriculture
Chapter 10 Ozone Damage to U.S. Agriculture
Chapter 11 National Freshwater Recreation Benefits of Water
Pollution Control
Introduction
Discharge Reductions and Linkages to Ambient
Quality and Fish
Behavioral Economic Aspects of the Study
Chapter 12 A Survey Research Method for Estimating National
Water Quality Benefits
Introduction
Research Procedures
Water Pollution Ladder and Value Levels
Willingness to Pay Questions and Answers
Chapter 13 The Values of Visibility in the National Parks...
Chapter 14 Benefits from Controlling Acid Rain
Chapter 15 Benefits from Avoiding Groundwater Contamination.
Chapter 16 Concluding Notes
BIBLIOGRAPHY
Reports to EPA Upon Which the Volume is Based
A. Other Volumes in This Series
B. Subsequent Reports
Additional Publications from EPA Grants for
Improving Methods for Estimating Benefits from Air
Quality
Supplementary Readings
47
47
53
55
58
59
64
70
70
71
73
77
77
78
78
81
84
90
93
98
102
102
102
103
105
107
VI
-------
VII
-------
PART I: BASICS
1
-------
1
CHAPTER 1
INTRODUCTION TO BENEFIT-COST ANALYSIS
In the 1960s, the people of the United States became increasingly
aware that the fruits of economic development were infected by the rot
of environmental deterioration. Late in the decade and early in the
1970s, concern grew to such an extent that a number of laws were passed
by the Congress aimed at not only stemming the deterioration of the
environment, but improving its quality as well. As we move into the
1980s, environmental concerns, as attested by public opinion polls, are
still vividly alive, but are being increasingly balanced by economic
considerations. In this atmosphere, there has been heightened interest
in the question of whether the costly environmental regulations that
have been put in place are, in fact, worthwhile. To try to shed
some light on this question, appeal is often made to an economic
evaluation method called benefit-cost analysis.
Benefit-cost analysis was developed initially to evaluate water
resources investments by the federal water agencies in the United
States, principally the United States Bureau of Reclamation and the
United States Corps of Engineers. The general objective of the method in
this application was to provide a useful picture of the costs and gains
associated with investments in water development projects. The
intellectual "father" of benefit-cost analysis was the nineteenth
century Frenchman, Jules Dupuit, who in 1844 wrote an often cited study
"On the Measure of the Utility of Public Works." In this remarkable
article, he recognized the concept of consumers' surplus (which is
explained in the next chapter) and saw that as a result, the benefits of
public works usually are not the same thing as the direct revenues that
the public works projects will generate.
In the United States, the first contributions to development of
benefit-cost analysis did not come from the academic or research
communities, but rather from government agencies. Water resources
development officials and agencies in this country have from the very
beginning of the nation been aware of the need for economic evaluation
of public works projects. In 1808, Albert Gallatin, President
Jefferson's Secretary of the Treasury, produced a report on
transportation programs for the new nation in which he stressed the need
for comparing the benefits with the costs of proposed water
improvements. Later the Federal Reclamation Act of 1902, which created
the Bureau of Reclamation and was aimed at opening western lands to
irrigation, required economic analysis of projects. The Flood Control
Act
2
-------
3
-------
of 1936 proposed a feasibility test for flood control projects which
requires that the benefits "to whomsoever they accrue" must exceed
costs.
In 1946, the Federal Interagency River Basin Committee appointed a
subcommittee on benefits and costs to coordinate the practices of
federal agencies in making benefit-cost analysis. In 1950, the
subcommittee issued a landmark report entitled "Proposed Practices for
Economic Analysis of River Basin Projects." This document was fondly
known by a generation of water, project analysts as the "Green Book."
While never fully accepted either by the parent committee or the
pertinent federal agencies, this report was remarkably sophisticated in
its use of economic analysis and laid an intellectual foundation for
research and debate in the water resources area which made it unique
among other major reports in the realm of public expenditures. It
also provided general guidance for the routine development of benefit-
cost analysis of water projects which persists until now, even though a
successor report does presently exist which is more adapted to the
conditions of the present day.
Following the "Green Book" came some outstanding publications from
the research and academic communities. Several volumes which appeared
over the past two-and-a-half decades have gone much further than ever
before in clarifying the basic ideas underlying benefit-cost analysis
and the methods for quantifying them. Otto Eckstein's Water Resource
Development: The Economics of Project Evaluation (Harvard University
Press), which appeared in 1958, is particularly outstanding for its
careful review and critique of federal agency practice with respect to
benefit-cost analysis. A clear exposition of principles together with
applications to several important cases was prepared by Jack
Hirshleifer, James DeHaven, and Jerome W. Milliman in Water Supply:
Economics and Policy (University of Chicago Press, 1960). A later study
which was especially notable for its deep probing into applications of
systems analysis and computer technology within the framework of
benefit-cost analysis was produced by a group of economists, engineers,
and hydrologists at Harvard and published under the title Design of
Water Resource Systems in 1962 (Harvard University Press). The
intervening years have seen considerable further work on the technique
and a gradual expansion of it to areas outside the water resources
field, some of them more or less natural extensions of the work on water
resources. For example, the last two decades have seen many attempts to
evaluate the benefits of outdoor recreation--both water-related and
otherwise. A relatively recent book which looks at some applications
other than water-related ones, but which is in the mainline of the
traditional benefit-cost analysis, is Ezra Mishan, Cost-Benefit Analysis
(Praeger Publishers, 1976).
But the most striking development in benefit-cost analysis in
recent years has been its application to the economic and environmental
consequences of new technologies and scientific and regulatory programs.
For example, the Atomic Energy Commission (before the Energy Resources
and Development Administration and then the Department of Energy were
created) used the technique to evaluate the fast breeder reactor
program. A report on this study is found in U.S. Atomic Energy
Commission, Division of Reactor Development and Technology, Updated
(1970) Cost-Benefit Analysis of
4
-------
the U.S. Breeder Reactor Program, Washington 1184 (January 1972).
The technique has also been applied to other potential sources of
environmental pollution and hazard. Two studies which come to quite
contrary conclusions have been made of the Automotive Emissions Control.
Volume 4, The Costs of Benefits of Automotive Emissions Control Series
No. 19-24, Washington GPO (September 1974) was prepared by a committee
of the National Academy of Sciences. The other study from a major
automotive producer is reported in Clement J. Jackson, et al., "Benefit-
Cost Analysis of Automotive Emissions Reductions," Research Laboratory,
General Motors Corporation, Warren, Michigan, CMR 2265 (October 15,
1976) . Other studies have been or are being conducted in the area of
water quality improvement policies, emissions control from stationary
and mobile air pollution sources, and regulation of toxic substances.
Even while the technique was limited largely to the relatively
straightforward problem of evaluating public works, there was much
debate among the economists about appropriate underlying concepts and
methods of making quantitative estimates of benefits and costs--
especially of benefits. Some of the discussion surrounded primarily
technical issues, e.g., ways of computing consumer surplus (the idea
referred to earlier and explained later) and how best to estimate demand
functions (also explained later) for various outputs of projects.
Others were more clearly value and equity issues, e.g., whether the
distribution of benefits and costs among individuals or regions needed
to be accounted for or whether it was proper to consider only the sums
over all affected parties. Another central issue was what the proper
weighting of benefits and costs occurring at different points in time
was to be. This is known as the "discounting" issue. The term refers to
the question of how to take into account the fact that normally the
further into the future gains or losses accrue, the less heavily they
are weighted by those who stand to do the gaining or losing.
Application of benefit-cost analysis to issues such as nuclear
radiation, the storage of atomic waste, and the regulation of toxic
substances in the various environmental media (both those substances
which are immediately toxic to man and those which affect his life
support or value systems) aggravate both the conceptual and
quantification problems which existed in water resource applications.
There are several reasons for this.
Firstly, while water resource applications often involved the
evaluation of public goods (in the technical economic sense which is
explained in the next chapter) , the bulk of outputs from such projects
are irrigation water, navigation enhancement, flood control, and
municipal and industrial water supplies. These outputs can usually
be reasonably evaluated on the basis of some type of market price
information because often private developments produce similar or
closely related outputs. In the new application, we are dealing
entirely with situations in which useful information from existing
markets is difficult, if not impossible, to establish.
Secondly, such matters as nuclear radiation and toxic materials
relate to exposure of the whole population or large subpopulations to
very subtle
influences of which they may be entirely unaware. It is difficult to
know what normative value individual preferences have under these
5
-------
circumstances, and clever methods of quantifying damages (negative
benefits) have to be evolved.
Thirdly, the distributional issues involved in these applications
concern not only monetary benefits and costs, but the distribution of
actual physical hazard. For example, residents of an industrial city may
suffer ill health resulting from pollution associated with the
production of goods consumed in another locality. While it is not out
of the question that monetary equivalents to these risks could be
developed, the ethical value issues involved appear to be deeper than
just the associated economic returns. This is especially so if
compensation is not actually paid to damaged parties as in practice it
is usually not.
Fourthly, we are in some cases dealing with long-lived effects of a
policy decision which could extend to hundreds of thousands of years and
many, many human generations. This situation raises the question of how
the rights and preferences of future generations can be represented in
this decision process. Realistically, the preferences of the existing
generation must govern. The question is whether the simple direct
desires of existing persons are to count exclusively or whether justice
demands that the present generation adopt some ethical rule or rules of
a constitutional nature in considering questions of future generations.
Thus the new application of benefit-cost analysis bristle with
ethical, value, and quantification issues. A group of researchers
located principally at Resources for the Future and the Universities of
Wyoming, New Mexico, and Chicago have, for the last several years, been
working on a research program aimed at making progress in the basic
understanding and analysis of these issues. In the present book, a
nontechnical summary of results from one of the most substantial thrusts
of this research--methods development and quantitative estimation of
benefits from air and water pollution control (air and water quality
maintenance or improvement) is presented. This program of research has
received sustained support from the U.S. Environmental Protection
Agency. A person wishing to study the details and technicalities of the
research studies underlying this brief nontechnical volume describing
these studies is referred to the bibliography at the end of this book.
For simplicity, references are held to a minimum in the exposition
itself.
Before proceeding specifically to a discussion of the methods and
results of the research, it will be useful to describe, in general
terms, some of the basic ideas from the discipline of economics which
were central to this research enterprise. Also the next few chapters
display its inherently interdisciplinary character.
But before doing so, I wish to underline what this book is and
what it is not. It is not an effort to provide a comprehensive review of
environmental benefits studies in general. The case material in it comes
from EPA-sponsored studies of air quality and water quality, conducted
in a coordinated way over the course of a number of years primarily at
the
6
-------
University of Wyoming, the University of New Mexico, and Resources for
the Future. While this therefore is not an entirely comprehensive review
of research in the area, some bounds had to be set, and the one chosen
seems reasonable on three grounds: (1) I have had some personal
involvement in nearly all of the projects discussed and therefore feel
more qualified to write about them than if I had only read about them;
(2) these projects span the range of methodologies that have been
developed for benefits assessment work including bidding games, surveys,
property value studies, wage differentials, risk reduction evaluation,
and mortality and morbidity cost estimation (all of these will be
explained subsequently); and (3) they represent the results of a
reasonably coherently planned program of research. Accordingly, the
book contains a relatively complete picture of the state of the art of
benefits measurement for environmental improvements as of 1983.
However, a further point should be made, and that is that these studies
are deliberately at the frontiers of the benefit measurement craft.
Their chief intent was methodological improvement, and the reader should
give primary attention to that aspect. Quantitative estimates of
benefits are given but they should be regarded as preliminary and
experimental in character, and at best an order of magnitude indication
of the actual numbers. For this reason, I have not adjusted results for
inflation even though they accrued over several years. They are in the
dollars of the late seventies and early eighties. To refine them further
would confer on them an unfounded aura of accuracy.
7
-------
CHAPTER 2
WHAT ARE ECONOMIC BENEFITS
INTRODUCTION
While this book is intended to be a nontechnical presentation of
our research on air and water quality benefits, some knowledge about a
few key concepts from economic theory is essential to understanding both
the research approaches taken and the results attained. The most
central of these concepts is that of an economic demand for a good (a
material object which is valued by people), or for a service.
When economists speak of demand, they are referring to the relationship
between the real or hypothetical price of a good or service and the
amount of it consumers actually buy or would wish to buy per unit time
at that price. Except in very unusual cases, one of which actually
occurs in chapter 13, the amount consumers will want to take will be
less the higher the price. The discussion here of economic demand is
simple and straightforward, but very compact.
It is important to keep one distinction clearly in mind when
discussing economic demand. That is, the distinction between the demand
of one individual or household--individual demand--and the "added up"
demand of all individuals or households demanding that good or Service--
aggregate demand. The latter is sought in doing benefit analysis but it
is logically derived from the former.
INDIVIDUAL DEMAND
Let us start with a look at individual demand. Consider the
following numerical example of an individual's price quantity
relationship for the fictitious commodity widgets.
At a price of eight dollars, the consumer will buy no widgets, at
six dollars, he will buy two, and so on. If, for whatever number he
does wind up buying, he is charged the same amount for each one (this is
the usual practice in actually existing markets) then the third column,
in which the price is multiplied by the number taken, will indicate how
much he actually does pay. But if we could figure out a way to make him
pay the maximum he
1 . There is also the pertinent concept of derived demand.
On the theory that "enough is too much," we will postpone discussion of
this idea until we need it in connection with the case study presented
in chapter 8.
8
-------
Price
of
Widgets
Quantity
Taken by
Consumer
Price
Times
Quantity
Price Times
Incremental
Quantity
Accumulated
Price Times
Incremental
Quantity
8
0
$ o
0
$
0
7
1
7
7
7
(7)
6
2
12
6
13
(7 + 6)
5
3
15
5
18
(7+6+5)
4
4
16
4
22
(etc.)
3
5
15
3
25
2
6
12
2
27
1
7
1
1
28
0
8
0
0
28
is willing to pay for each individual unit (column four) or be deprived of having
any widgets at all, then the accumulated price times incremental quantity shown in
the last column would reflect his total willingness to pay for widgets. This is the
amount he would pay in an "all or nothing" ' situation where he either pays
everything he would be willing to pay or he is deprived of widgets altogether.
Now suppose that our consumer decides he wishes to buy 5 widgets because the
going price for widgets is $3 per item. He then actually pays $15, but if he had no
alternative but to pay the maximum he would have been willing to pay, then he would
have paid $25 for the three. The difference between what he did pay and what he
would have been willing to pay, $10, may be thought to be some extra benefit which
the consumer gets because there are such things as widgets available in the market.
But because they are uniformly priced at a level less than his maximum willingness
to pay, he gets this extra benefit. This additional value is called consumer's
surplus by economists. If it were to be the case that the consumer is not required
to pay anything for the widgets, he takes eight and his consumer's surplus will be
equal to his total willingness to pay--$28. In all cases where there is a positive
price, his total willingness to pay will be greater than what he actually does pay
because it will include what he actually does pay and his consumer's surplus. For
example, if he buys four widgets his willingness to pay equals what he actually does
pay plus his consumer's surplus (i.e., $16 + $6).
It is usual in expositions of consumer demand theory to express these ideas
graphically by plotting a demand curve for the individual. Below is a plot of the
numerical example just reviewed. In the simple example, the
9
-------
Price 4 ¦
0
2
1
12 3 4 5 6 7 8
Quantity
demand curve is a straight line. We generate this line by plotting a price quantity
pair point for each of the pairs shown in the numerical example, with interpolation
between the points. It is pretty apparent that the accumulated price times
incremental quantity column (willingness to pay) is the accumulated area under the
demand curve. To see this, observe that every individual price times quantity pair
makes a box on the graph as is shown more abstractly below.
Since the curve represents every possible combination of such Ps and Qs (all
possible boxes--imagine their width to be vanishingly small), it follows that the
area under the whole curve is equal to the consumer's willingness to pay at zero
price for Q.
Again, then, more abstractly than in the numerical example above, let us use a
graph to review all the main ideas we have defined so far.
P x Q = area of this box, e.g.,
amount actually paid
for this Q
P
Price
Q Quantity
10
-------
11
-------
actual price
of the good
Demand curve
(a)
consumer
surplus
(b) P x Q
amount actu-
ally paid by
consumer
Area (a) + (b) + (c) = total
willingness to pay for the
good at zero price
Area (a) + (b) =
(c) willingness to pay
for quantity Q
Q
+
Quantity
Quantity taken at
that price
AGGREGATE DEMAND
So much for the individual consumer. But for many purposes (some of which will
become clear later) we are interested in the total demand by all consumers for a
good or service (in this case, widgets). How, then, do we add up the demands of all
consumers in this market? If we are willing to make the assumption that all persons
in the market for widgets should be treated equally, that is to say, everyone's
demand counts the same in making up the sum, the answer is very easy--we just add up
the quantities demanded at every price. For example, let us assume that there are
two individuals in the widget market and both are just alike--let's say both are
like the one in the numerical example. In this case, the aggregate demand would be
just double the individual demand at any given price. For example, at the price of
$5, aggregate Q would be 6, P x incremental Q would be $30, and P x Q accumulated
would be $3 6.
Again I illustrate this adding up process a little more abstractly and
generally with a graph.
Distance A
Distance B
ft T1 <"¦» A
ii + D = HA
ZQ = 2A
Price
Demand curve of individual 1
+¦ Demand curve of individuals 1 am
2 added together where indivi-
dual 2's demand curve is
similar to individual 1's
Quantity
12
-------
There is no reason why individuals would need to be similar to
make the adding up work. Everything is done the same way if they are
not, only the numbers are different. Once an aggregate demand curve has
been calculated, the concepts of willingness to pay and consumer's
surplus apply to it in the same way as to the individual demand curve
(still assuming we are willing to treat everyone equally for this
purpose).
Stated in its broadest terms, the objective of the research
described in this document is to develop methods to derive estimates of
the demand (willingness to pay) for cleaner air and water which would
then be at least loosely comparable to the demand for other goods and
services. This is to permit, at least roughly because of the
uncertainties involved, comparison of the value consumers place on
cleaner air relative to other goods and services they buy. In practice,
this is a very hard problem. But, unfortunately, even from the
standpoint of ideas and concepts, we are not yet ready to proceed to
quantitative economic analysis. In fact, cleaner air or water are not
goods similar to widgets or the many real goods and services, ranging
from houses to pins, that can readily be bought and sold in markets.
Economists refer to goods like widgets as private goods, and goods like
cleaner environments as public goods.
PRIVATE GOODS AND PUBLIC GOODS
In the economist's lexicon, widgets are private goods because they
are divisible and separable. If you buy a widget and use it, that same
widget does not at the same time render a service to me. If I
buy and eat a banana, you cannot buy and eat that same banana. Such
goods are easy for the private sector to produce and market because they
come in distinct, divisible units and can be sold to distinct, divisible
buyers. Should you, however, go and buy cleaner air, for example, in
the city where you and I reside, say by paying industries to clean up,
the services of that cleaner air are at the same time available to me,
even though I didn't pay anything for them. Such goods are called
public goods because their units are not divisible and distinct. Their
services are available to many persons at the same time, including those
who don't pay for them, and unlike private goods the use of their
services by one person does not diminish their availability to others.
Private markets are very bad at producing such goods; indeed, there
usually is no private economic incentive to produce them at all because
while many people could benefit from them, no single individual has a
sufficient incentive to pay for them.
Two chief implications for the research reported in this book flow
from this situation. First, while in principle it is possible to think
of an individual demand curve for cleaner air or water just like a
demand curve for widgets, there usually will not be market price
information which will help directly in defining such a curve.
Sometimes, as we will see further on, such information is helpful
indirectly. This means further that development of methods for
obtaining information on how consumers value or would value cleaner air
or water if they had more information, is a very important and
difficult task. To develop such methods was, as already stated, the
chief objective of our research.
13
-------
A second implication is that even if we have individual demand
curves for public goods, we cannot properly add them up in just the same
way as for a private good. The way we added up for the private goods is
called summing horizontally. Individual demands for public goods must
be summed vertically.
To see this, refer back to our widgets example. Assume that
instead of demand for widgets, the columns refer to successively lower
prices for air quality improvements for an individual consumer and the
quantities of improvement the consumer would want at those prices.
P x Q and P x Q accumulated have the same interpretation as for private
goods for this one individual. But now let us add a second consumer
as we did in the private goods case. With the second consumer added in,
it does not mean that more units of quantity of cleaner air will be
taken at a given price, as was the case with the private good. The
same units of quantity are available to both consumers. Thus, the
willingness to pay for up to three units of cleaner air is $18 for the
first individual plus $18 for those same three units, or a total of $36.
As noted, the kind of summing done here is called vertical summing in
contrast to the horizontal summing for private goods. Again, we can
illustrate this graphically. It is easier to show the procedure when
demand curves for the two individuals are not equal, so our illustration
assumes they are not. In the graph below, individual
demand curves are designated D1 and D2 ' For any given level of air
quality, say Q, the willingness to pay for that level (the cross-hatched
area) is the willingness to pay of. D, plus the willingness to pay of
D2 for the same quantity of air quality improvement.
This total willingness to pay for q units of clean air is in
economic terminology the "benefit" of q units of clean air. Since no
price is charged for these q units it is also the consumers surplus
associated with the provision of q units of clean air.
Price
Q
Quantity
14
-------
COMPENSATION
A final note on concepts of demand; economic reasoning indicates
that when we are considering a situation in which persons are deprived
of something they otherwise would have had, as when previously clean air
is polluted, willingness to pay for the clean air is not the fundamental
test of its value to them. Rather, if they are to be as well off as
before the change, one must ask how much they would have had to be
compensated to be as well off as before. Generally speaking,
willingness to pay is easier (although usually not easy) to estimate
than required compensation. Economic theory indicates that the
former will be equal to or smaller than the latter. In most of what
follows, we will concentrate on willingness to pay as a conservative and
usually more measurable quantity.
The aspiration in the quantitative studies reported here was to
estimate willingness to pay, but we shall see that we must often be
satisfied with results that resemble more the price times quantity
value. But we do have the advantage of knowing in which direction the
error lies in such an instance. We know from the earlier discussion
that P x Q will never be larger than willingness to pay and that usually
it will be smaller.
This completes the general discussion of economic benefits.
More specific topics in the area will arise in connection with the case
studies presented in Part II of this volume.
The next chapter in this part treats briefly an essential element
in the complete analysis of benefits from air and water quality
improvement. This is the matter of establishing the link between a
change in emissions of pollutants to the atmosphere or a water course on
the one hand, and the ambient environmental conditions on the other,
which, in some manner, adversely affect human beings.
There are two steps or linkages in the analysis of benefits from
reducing emissions, the one just mentioned, and, the other, once that
link of emissions to human effect is established, what economic value is
to be placed on that effect. The latter is the subject, in abstract
terms, of chapter 4. The case studies presented in the following
chapters concentrate primarily on quantifying the value to be placed on
various pollution effects. But in the next chapter, as mentioned, I
discuss the linkages between emissions and effects on humans.
While this is not strictly an economic problem, the economist
endeavoring to estimate benefits must often study these linkages as well
because there is in many cases no pre-existing information about them.
15
-------
CHAPTER 3
LINKS BETWEEN ACTIONS THAT AFFECT THE
ENVIRONMENT AND EFFECT ON HUMANS
INTRODUCTION
As just stated, an essential element in estimating the benefits
from air and water pollution control programs is an understanding of how
emissions control affects the environmental conditions which humans
value. Such effects can be rather direct and easy to perceive, as when
visibility is impaired, or quite indirect and difficult to perceive, as
when air pollution produces chronic illness or when agricultural
productivity is reduced by air contamination. This chapter briefly
discusses the various linkages, and methods of estimating them, between
emissions and ambient conditions that directly or indirectly affect
humans. Understanding these linkages is central to the various
illustrative cases of economic evaluation discussed later.
HEALTH LINKAGES
Concern about health effects has been the basis for most of our
air pollution legislation in the United States. Linkages between
emissions and health are subtle and difficult to establish, especially
when one wishes to link changes in emissions to health states (as is
necessary if we wish to estimate the demand for improved air quality or
the demand for air quality maintenance). Four kinds of information are
needed: emissions, translation of those emissions to concentrations in
the environment, dose-response relationships (i.e., how are specific
concentrations of an air pollutant related to health), and the
population at risk. The latter three of these items are hard to
estimate. Translations of emissions into concentrations in the
environment (e.g., tons of sulfur oxide emitted into parts per million
of sulfates at various points in the surrounding air) is best
accomplished by means of special computer models called dispersion
models. These are imperfect at best, and it is usually not possible
to verify them against observed conditions. The linkage between
concentration in the environment and health effects is also hard to
establish, especially if we are concerned with chronic as opposed to
acute effects.
For example, there has been much concern about possible links of
air pollution to cancer. But cancer is a disease that usually appears
many years after there has been exposure to carcinogenic substances--
often fifteen or twenty years later. Therefore, it is very hard to sort
out Possible causes. Essentially, there are two ways of trying to make
the
16
-------
link--epidemiology and animal studies with subsequent extrapolation to
human beings. The latter have well-known deficiencies, especially
with respect to something as complex and subtle as air pollution, and we
made only limited use of them in our research. Epidemiology observes
conditions as they exist in actual environment, and, trying to control
for other factors which might be related to health, endeavors to isolate
the effect of air pollution. Our effort to develop improved methods for
quantifying the benefits of air quality improvement involved several
epidemiological experiments which are reported in connection with cases
reviewed in Part II. Almost needless to say, efforts to establish the
link between air pollution and health are afflicted by great
uncertainty.
VISUAL QUALITY
An emerging air quality issue of central importance, especially in
the West, is the impairment of visibility due to air quality
deterioration. In this case, the linkage between emissions and effects
on humans is by direct perception of the degraded conditions. However,
we are not interested only in how some fixed condition is perceived,
but, in accordance with our discussion of economic demand in the
previous chapter, we wish to know what persons would be willing to pay
for alternative, increasingly better, levels of air quality. Therefore,
even though a person can perceive conditions directly, we must find ways
of simulating situations other than those that exist at any given time.
While I reserve deeper discussion of how values can be attached to such
conditions until the next chapter, we can say here that the main method
employed is asking people, with carefully structured questions, how much
they would be willing to pay for improved conditions. To solicit such
information, simulated conditions are presented in visual form to the
interviewee. Generally, this is done by means of photographs which have
been taken during actual episodes of clean and dirty air that, at one
point in time or another, have actually existed in the particular area
being viewed. A technique which potentially is an advancement over this
procedure is computer simulation of changed conditions. In this
technology, a slide of a particular scene is put in digital (numerical)
form so that it can be replicated by a computer on a high resolution
television screen. Then computations are made about the effect
which a hypothetical change in emissions, associated, say, with a
projected new power plant at some specified distance and direction from
the scene, would have on visibility. Since this calculation is also
numerical, the computer can then simulate in pictorial form, on the
television screen, the changed conditions of visibility. Development of
this latter technique has been part of the projects reported here
(although not funded by EPA), but efforts made to get consumer values
for visibility, because of the then existing state of the art, had to be
based on actual pictures.
LINKS TO AGRICULTURAL PRODUCTIVITY
Agriculture may be adversely affected by air pollution.
Plants may suffer from leaf burn due to acid rain resulting from sulfur
or nitrogen compounds emissions or may be weakened and made more subject
to disease by
17
-------
exposure to ozone. Since vegetation may be influenced by many factors,
only one of which is pollution, isolating this effect is not so
straight-forward as it might appear. However, that such damage does
exist is well documented. Associations between monitored levels of air
pollution and crop production are reasonable well established,
especially in Southern California, the locale for the study reviewed in
chapter 9. But as in the case of visual impacts and health, we are
interested not only in what effect existing levels of pollution have on
production, but also in what impact changes in pollution levels would
have. This once again means that, at least in principle, estimates of
real or hypothetical emissions changes must be translated into ambient
conditions with dispersion models, crops that may be especially
sensitive identified, and exposed acreage calculated. Estimating the
effects on consumer welfare via their derived demand (explained in
chapter 9) for cleaner air for agriculture presents a particularly
subtle and difficult problem. But that is a subject for the next
chapter.
LINKS IN WATERCOURSES
I have repeatedly emphasized the need for having a linkage between
pollution discharge and effects on things in the environment that man
values or on man himself. Unfortunately, in the air, the needed
dispersion models are only available in some places--no nationwide model
is available or, for that matter, feasible in the present state of the
art. Accordingly, often various simplifications must be made in
actually doing air quality benefits studies--especially ones that are
aimed at estimating national benefits. These short cuts will be
described in connection with the cases as the need arises. Fortunately
we are in somewhat better shape in the water quality area where impacts
of changes in effluent discharges on the aquatic environment must
similarly be forecast. Resources for the Future has built, maintains,
and is steadily improving a National Water Quality Network Model.
Constructing the model was not part of the EPA-sponsored research, but
its results were incorporated into that research. This model simulates
in a computer water quality changes associated with changes in effluent
discharge in the main water courses of the nation as shown in the map
below.
The network of water bodies contains 3 04 rivers, 17 5 lakes and
reservoirs, 37 bays, 10 segments of Great Lakes shorelines, and 26 ocean
shoreline segments. Pollutants can be injected into the system at the
nodes (municipal and industrial discharges) and uniformly between them
(nonurban runoff). The computer model then simulates the transport,
degradation, and transformation processes that occur in the water body
and calculates a number of water quality characteristics at any point in
the system taking account of all of the points of discharge that affect
that location. This capability, when translated further into areas of
water rendered suitable for various recreational activities by pollution
control policies, proved very useful in the benefits from recreational
fishing project described in chapter 12. Unfortunately the model
presently can handle only a few of the more conventional better
understood types of pollution--biochemical oxygen demand and suspended
sediment, for example.
18
-------
KBY:
Map of rivets, lakes, and reservoirs included in the National Water Quality Network model.
Subtler influences on water quality, such as the effects on aquatic
ecosystems of the introduction of acid from environmental sources
presently elude it.
AQUATIC ECOSYSTEM LINKAGES
Over time, it has become increasingly apparent that rainout and
other types of deposition of materials from the atmosphere are major
sources of contamination of water courses. Special interest and concern
has come to focus on acid deposition. When fossil fuels, especially
coal, are burned, compounds of sulfur and nitrogen are released along
with the other flue gases. Automobiles are also an important source of
nitrogen emissions. Through chemical transformation processes in the
atmosphere, these substances are partly converted to sulfuric and nitric
acid. When this acid rains out of the atmosphere or is otherwise
deposited in water courses, especially lakes, they may become so acid
that they cannot continue to support fish life. Also, increasingly acid
soils can affect plant life adversely. Understanding the link between
emissions at particular sources and such ecological effects is
difficult, and research on the question is in its infancy. In
principle, we need again to understand quantitatively the processes of
dispersion in the atmosphere (in this case, for very long distances--
possibly thousands of miles) , deposition processes, effects on acidity
of the stream and related
19
-------
phenomena (for example, increased acidity may cause toxic heavy metals
to dissolve and become a problem) , and finally, the ways in which
aquatic life is affected by the acidity. In practice, as we will see in
chapter 14, we must make do with much less knowledge than this in our
quest for the benefits of controlling acids from the atmosphere.
Moreover, what we do know about the linkage between increasing acidity
and fish life suggests that it is quite complex. For instance, it seems
that as a certain critical level of acid in the water body is reached,
damage to aquatic life mounts drastically with small further increases,
but that damage then increases much more slowly, if at all, with further
increases. Also, once damage has occurred, it may not be reversible by
any practically available technology. Both these characteristics have
substantial implications, as we will see in chapter 14, for the economic
evaluation of benefits from controlling acidity in water bodies.
MATERIALS DAMAGE
As well as having adverse aquatic ecosystem effects, the
deposition of acid or its precursors is the major cause of materials
damage from air pollution. Again, dispersion and deposition processes
must be understood, but the actual damaging effects are chemical rather
than biological in nature. For example, sulfuric acid reacts with the
carbonate in limestone and destroys the stone. In addition, acids etch
metals and cause corrosion. Similarly, fabrics and plastics can be
damaged. Unfortunately, quantitative understanding and predictability
of these phenomena is extremely primitive so that, once again, radical
assumptions must be employed if damages, especially damages associated
with changed conditions, are to be estimated.
GROUNDWATER LINKAGES
One of the most difficult to simulate linkages between pollution
discharge and changed conditions in the environment is in the case of
groundwater. This is so because (1) far fewer resources have gone into
developing such an understanding than is true of surface waters and the
air, (2) ground- water flow is often highly complicated, and (3) it is
very difficult and expensive to make measurements (holes must be bored).
Because it is a highly specialized area and because establishing the
needed linkages was such an integral part of the benefits study of
controlling groundwater contamination, further discussion will be
deferred until I turn directly to that study in chapter 15.
The discussion of the present chapter, unfortunately, illustrates
that, even though it is in the domain of "hard" sciences, understanding
of exactly how natural systems are affected by man's discharge of
polluting substances is still very limited. The uncertainties of
knowledge about these linkages are fully as great as the uncertainties
about how to do the actual economic evaluations. Thus we are
studying, in the experiments reported in this volume, something more
akin to a craft than an exact science.
20
-------
The next chapter discusses, again in very general terms, some
methodological aspects of placing an economic value on air and water
pollution effects on the environment once they are identified and
quantitatively estimated. Each of the methods discussed is employed and
further explained in one or more of the case studies in Part II.
21
-------
CHAPTER 4
PROBLEMS OF ASSIGNING ECONOMIC VALUES
INTRODUCTION
Once links have been established between humanly controllable
actions that affect the environment and the associated direct and
indirect effects on humans, then the central problem addressed by the
research reported here arises--how to measure the economic demand for
cleaner air and water. That is to say, what is the economic value to be
attached to a given or successively higher levels of improvements to air
or water quality or to protect the existing level of quality from
deterioration? The methods used to make estimates of these values
necessarily differ as among the different types of effects associated
with air and water quality deterioration. This is partly inherent in
the different situations, for example, whether the effect is directly or
indirectly on consumers, and partly a matter of the types of data it is
practical to acquire. As further background for discussion of case
studies in the following chapters, I briefly review some central issues
in the economic evaluation of cleaner air and water.
VALUING RISK TO HEALTH
The studies of health effects reported in the next two chapters
focus on the possibility that air pollution and contamination of
groundwater may cause chronic disease, which in turn may contribute to
higher death rates (mortality) or nonfatal sickness (morbidity). One
central question, if one is to calculate a benefit in monetary terms, is
what value to place on reduced mortality. How much would people be
willing to pay for a reduction in their risk of earlier death or how
much would they have to be compensated to voluntarily accept an increase
in this risk?
Economists in the past have attempted to value human life as the
future earnings over an individual's lifetime. This approach, however,
is now no longer viewed as acceptable. In the first place, it assumes
that the value of life can, in fact, be measured in economic terms--a
point certainly open to debate. Second, it implies that the lives of
children, housewives, retired, and other unemployed individuals are
worth less than
2. For those familiar with the concept of present value, it
should be explained that the value used is actually the discounted
present value of expected future earnings.
22
-------
the lives of employed heads of households. Nearly everyone would find
these implications ethically unacceptable.
An American economist named Thomas C. Shelling was, nearly twenty
years ago, apparently the first to distinguish between the concept of
the cost of statistical risk and efforts to value human life based on
lost earnings. The cost of risk idea is ethically more appealing than
attempts to value a particular human life. The effort here is to put a
value on a small increase or decrease in the probability of death for
anonymous, statistical persons. Implementation of this approach has
usually involved a search for information about how much people have to
be compensated to voluntarily accept a small increase in risk in
occupations differing in riskiness--say the risk of additional death per
thousand persons. Thaler and Rosen (N.E. Terleckyj, ed., Household
Production and Consumption, 1976, Columbia University Press) , using wage
differences between jobs varying in the level of job-associated risk of
death, were apparently the first to estimate explicitly the value of
changes in safety. They observed that workers in high risk jobs receive
higher wages, and a value of safety can be imputed by examining these
risk-related wage differences. Other factors that influence wages were
statistically held constant by use of a technique called regression
analysis (this method is briefly explained in the following chapter).
Unfortunately, however, the Thaler and Rosen study dealt with a class of
individuals who, because they are engaged in risky occupations may be
more willing to accept risk than the rest of the population. Even so,
the estimate they make suggests that a small reduction in risk over a
large number of individuals which saves one life is worth about $340,000
(in mid 1970's dollars) . This is far higher than the numbers obtained
in lost earnings studies. Another study, Blomquist (Journal of
Political Economy, June 1977), which examines seat belt use, suggests
that the figure for a lost life might be $260,000. This study first
estimates how people value their own time and then imputes a value of
safety from the amount of time a sample of individuals spent in buckling
up seat belts. It may be noted that unlike the Thaler and Rosen result,
this is a "willingness to pay" rather than a "compensation" measure.
The result may, however, also be biased downward because individuals
seem to perceive risks differently when an element of personal control,
such as driving an automobile, exists rather than when an involuntary,
individually uncontrollable risk is at issue, as is the case with
environmental risk. Finally, Smith (Law and Contemporary Problems,
Summer-Autumn 1974) in a study similar to Thaler and Rosen-s, but for a
more typical population, found that the needed compensation to save one
life may exceed $1,000,000. Numbers even higher than this have been
reported in the literature.
Clearly, the cost of risk is not precisely known, and perhaps will
never be, since attitudes--in particular, risk averseness--presumably
can change over time, between groups, and can even vary in different
situations. But, we at least have a range of values with which to make
order of magnitude estimates of the costs of environmental risks.
Likely values lie between a quarter of a million and a million dollars
per life, valued in mid-1970s dollars.
23
-------
There are some additional observations to be made about valuing
mortality risk by a particular number derived from observed behavior of
people concerning risks.
First, no distinction is made with respect to age, sex,
employment, or other personal variables. To paraphrase Gertrude Stein,
"a life is a life is a life." This seems ethically acceptable, but might
well be the subject for debate.
Second, this analysis does not give attention to the pains and
suffering associated with different causes of mortality or to the cost
of nonfatal diseases (morbidity).
Third, the value obtained from existing studies does not vary with
the degree of risk. To put the matter in terms of the discussion of
economic demand in chapter 2, this means that the demand curve for
mortality reduction vis-a-vis pollution looks like that depicted in the
following figure.
Value per
life
number of lives
saved because
degree of clean
air achieved
Number of lives saved
because of cleaner air
While as stated in chapter 2, one generally expects price to
decline as quantity increases, this constant value may be defensible
within the present context because in the case of air pollution we are,
at most, speaking about small changes in the general risk to health.
Over such a small range, it is not unreasonable to think that the value
of risk reduction would remain about constant.
MORBIDITY
Pollutants can, of course, do much harm to health without actually
killing. A number of studies have tried to evaluate this harm by
estimating the number of days lost from work because of such pollution
and then, to get an economic value, multiplying those numbers of days by
the average wage rate. This procedure is incomplete for several
reasons: it does not value the cost of sickness for persons who are not
in the labor force, i.e., it neglects the disutility of the sickness
itself. Also it does not recognize that people can protect themselves
to some extent, and
24
-------
at some cost (say by installing an air filter), against sickness.
Approaches that recognize and account for these factors are sorely
needed.
VISUAL PERCEPTION
Introduction
As pointed out in the last chapter, questions about the value of
visibility impacts have become highly significant in air quality policy,
especially as it applies to conditions in the mountainous West, where
unusually clean air and the associated large, bright landscapes are
highly prized by many people. The question of how to value such effects
is a very difficult one. In an urban area, one might consider using
differences in housing property values as an indication of aesthetic
values people attach to air clarity--this approach is discussed further
on. But in scenic rural areas such as national parks, this is clearly
not feasible. Thus, it was necessary to develop and use alternative
methods.
The method chosen for our research used questions posed to
recreationists and others affected by visibility impacts in an effort to
discover their preferences and values. In all cases studied, the
respondent was confronted with an image of possible changes in air
quality at a particular site, in the form of carefully prepared
photographs, and asked to state a value for it. The respondent was also
asked to reveal other pertinent characteristics about himself or
herself. This approach is referred to in the trade as a "bidding
game." Respondents can be queried as to willingness to pay for the
cleaner air conditions, minimum compensation to accept a change,
potential site or activity substitutions for the one in question,
income, age, sex, etc. As is explained in connection with discussions
of cases in Part II, responses to these types of questions can be used
to estimate demand curves for cleaner air.
The major concern in using bidding games, or other survey
questionnaire techniques (such as the one discussed with respect to
water quality in chapter 12), to construct demand curves is that the
reply to questions may be biased either because the interviewee wishes
to deceive or because of problems in the way questions are posed.
Possible biases which could well exist in theory have been a major
preoccupation of researchers pursuing the bidding game and other survey
approaches. The main types of bias which have been identified in our
work as possibilities are: (1) strategic bias, which means that the
respondent may attempt to influence the outcome or result by not
responding truthfully; (2) information bias, which is bias resulting
from lack of complete information on the part of the respondent; (3)
starting point bias, where the respondent may be influenced by the
opening bid which is usually suggested by the interviewer; and (4)
hypothetical bias, which could result from inability to confront the
respondent with an actual situation, for example, using a photograph
rather than an actual scene.
The bidding game and other survey techniques are sufficiently
central to the research in several of the case studies reported in the
following
25
-------
chapters, and possible biases in results are sufficiently important,
that they merit a bit of special attention.
Strategic Bias
Most economists have long supposed that direct revelation of
consumer preferences for public goods (defined in chapter 2) would be
impossible. In particular, the so-called "free-rider problem" would
arise because the public goods situation gives individuals incentives to
misstate their preferences. For example, if nearby residents were asked
how much they were willing to pay to clean up the air near a power
plant, and if they suspected that control costs would be borne by
consumers and owners elsewhere, local residents might well have an
incentive to greatly overstate their actual willingness to pay since
they would, in fact, not have to pay anything. On the other hand, if
residents believed that they would be taxed an amount equal to their own
individual willingness to pay, then a clear incentive would exist to
understate their own true value, since their individual bid would have a
negligible effect on the outcome in any case.
It is thus apparent that different techniques aimed at eliciting
willingness to pay may generate their own variety of bias. For example,
if respondents are told that the average of their bids to prevent
construction of a power plant near a national park will be used to set
an entrance fee to the park, those individuals who suspect their bid to
be greater than the average bid will have an incentive to overstate
their willingness to pay. They, in fact, in principle have an incentive
to raise the average bid as close as possible to their own true bid. In
other words, individuals will, again in principle, have incentives to
misstate their own preferences in an attempt to impose their true
preferences on others.
Information Bias
Since bidding games are hypothetical, answers obtained through
these surveys will not be based on information or perceptions as
complete as would apply if consumers based answers on real experiences
which, unfortunately, is usually not possible. Typically, consumers do
reevaluate actual decisions on the basis of experience. Thus,
a recreator might respond to a hypothetical decrease in air quality at
one location with a low bid, thinking that other nearby sites would make
good substitutes. However, in a real situation, the recreator might
have found that other sites involved more travel costs and were less
satisfactory than imagined. Clearly, then, the information presented
to the respondent in a questionnaire situation relating to substitution
possibilities and alternative costs may well bias the stated willingness
to pay. On the other hand, there may be no amount of verbally conveyed
or written information that can fully substitute for actual experience.
Starting Point Bias
Central to the bidding game approach are questions on willingness
to pay (and/or compensation) for hypothetical changes in air quality.
It may
26
-------
be that it is better to ask the interviewee a question with a "yes" or
"no" answer rather than a question requiring independent quantitative
estimation on his part. Assuming that yes/no responses are desirable,
it is necessary to suggest a starting bid or minimal level of
compensation. Here the potential bias arises because the interviewee's
final reply may be influenced by the opening bid. This possible bias
comes from at least two possible sources. First, the bid itself may
suggest to the respondent the approximate range of appropriate bids.
Accordingly, he may respond differently depending on the amount of the
starting bid. Second, if the respondent values time highly, he may
become "bored" or irritated with going through a lengthy bidding
process. In consequence, if the suggested starting bid is substantially
different from his actual willingness to pay, the bidding process may
yield inaccurate results. The effect of these two types of starting
point biases may substantially influence the accuracy of bidding game
valuation and therefore the usefulness of this approach for assessment
of references with respect to air pollution.
Hypothetical Bias
The bidding game requires suggesting, by way of pictures, a change
in air quality such that it is believable to the respondent and
accurately depicts a possible potential change. In addition, the change
must be fully understandable to the respondent, i.e., he must be able to
understand Most, if not all, of its ramifications for him. Finally,
he must believe that the change might occur and that his bid might have
an effect on both the possibility and magnitude of change in air
quality. If these conditions are not fulfilled, the hypothetical nature
of bidding approaches will make their application to air quality issues
dubious and may bias the respondent's answers up or down. However,
unlike other types of biases identified, it is extremely difficult to
measure the extent of hypothetical bias since it depends not only on how
well structured the interview is, but also on uncontrollable factors
such as attitudes, style of presentation by interviewer, the
recreationist's "mood," etc.
Conclusions About Bidding Game Bias
To test for the presence and importance of bias and to assist in
developing methods in controlling for it, the research team ran a number
of "experiments" using bidding games and surveys. The experiments show
that all forms of bias can definitely exist. But it appears that
problems of strategic, information, and starting point bias are all
surmountable with proper questionnaire design and statistical analysis.
This, plus the comparison with an alternative valuation method in
chapter 7, and a technique developed in the National Water Quality
Survey reported in chapter 12, suggest that well-designed survey
techniques can produce reasonably reliable information about the value
of air and water quality and other public goods.
27
-------
WATER-BASED RECREATION
Much work has been done by economists on the problem of evaluating
water-based recreation and aesthetic values, and many methods have been
applied to the problem. These include bidding games, other types of
surveys, inferences from the value of waterfront properties, and a
method based on travel costs to particular sites. In the two studies
reported in this volume, a survey method was used in one and the travel
cost method in the other. For present purposes we distinguish between
bidding games and surveys, as hinted in the previous section, even
though they both ask respondents questions about willingness to pay.
The bidding games reported in this volume all pertain to the evaluation
of quality changes at particular sites, and the sample population may or
may not be, but usually is not, randomly selected from the general
population. For example, if the technique involves interviews at the
site, the sample population consists of those who happen to be at the
site during the interviewing, and there is no reason to believe that
those questioned actually are a random representation of the population
at large. Surveys, as the term is used here, always choose their
respondents randomly from the national population. This is an important
feature for the study reported in chapter 12 because it was explicitly
designed to provide national benefit estimates, and randomness permits
an extrapolation of the sample results to the whole population by
statistically acceptable procedures. In addition, this study endeavored
to measure benefits of water quality improvement that may accrue to
people even though they may not be direct users of these water bodies.
These benefits are variously called nonuser, intrinsic, or existence
benefits. They are explained in later chapters.
The other benefits study reported in chapter 11 also had national
level benefits estimation as its objective, but was "site-specific" if
site is interpreted to be a rather large geographical area and focused
only on actual or potential users of water bodies for recreational
purposes, specifically, sport fishing. However, it used neither a
bidding game nor a survey method. Rather, it employed the travel cost
method to evaluate benefits to recreational fishing. This method was
developed at RFF many years ago and is a well-established technique of
recreational benefits evaluation that has been used many times by
economists, planners, and others to evaluate specific recreation sites.
The novelty of the study reported in chapter 11 is its ingenious
application of the methods to the particular problem of obtaining
a national benefit estimate for recreational fishing associated with
water quality improvement. Actual applications of the travel cost
technique are often quite complicated. Here I wish only to convey to
the reader the general concept of how the method is used to construct a
demand curve for a recreation site. The basic idea is that increased
access cost associated with user distance from a desirable recreation
site will tend to affect recreation visits in the same manner as an
increase in access cost resulting from a hypothetical rise in an
admission fee. If it were feasible to experiment with the fee,
hypothetically setting it from zero to increasingly higher levels, it
would of course be possible to define a relationship between demand and
price (a demand function as discussed in chapter 2) . The basic
principle of the procedure can be clarified by a simple numerical
example.
28
-------
Assume that we have divided the "market area" for a recreation
site into four zones at different distances from it, and we have the
information shown in the table about them.
Access
(travel)
Number
Visits per
Zone
Population
cost to
site
of visitors
1,0 00 population
1
1, 000
$1
500
500
2
2, 000
$2
400
200
3
3, 000
$3
300
100
4
4, 000
$4
0
0
Total visits at zero entrance fee
1,200
If there is no entrance fee, there will be 1,200 visitors (say per
year) and that gives us one point on the demand curve, that for a zero
price. This is shown on the diagram below. Now let us assume that an
Hypothetical
Admission
Charge
1 2 3 4 5 6 7 8 9 10 11 12
Number of Visitors (in hundreds)
entrance fee of $1.00 per visit is levied. This is taken to have the
same effect on visitation rates as a $1.00 difference in access cost
related to distance. Accordingly the visitation rate in zone one will
drop to that of zone 2 which has a $1.00 higher access cost. Therefore,
instead of a visitation rate of 500 persons per 1,000 persons from Zone
1, the rate will drop to 200. Since there are 1,000 persons in Zone
1, this means that there will be a total of 2 00 visitors from there.
Zone 2's visitation rate will drop to that of Zone 3 which has a $1 . 00
higher access cost than Zone 2, i.e., it will drop to 100 visits per
thousand population. Since Zone 2 has two thousand inhabitants, this
means a total of 200 visits from Zone 2. By the same reasoning, there
will be no visitors from Zone 3. Thus, at a
$3
$2
$1
29
-------
$1.00 admission fee, there will be 400 visitors--200 from Zone 1 and 200
from Zone 2 . This provides us with another point on the demand curve
as shown in the figure. Finally, at a $2.00 entrance fee there will
be no visitors from any of the zones. This produces still a third
point on the diagram--the point at which the quantity demanded will fall
to zero.
Obviously, this example is meant to be as simple as possible, and
because it established only a few points on the demand curve, would
provide only a very rough approximation of an actual curve. But the
principle is the same even in much more complicated applications.
VALUING AGRICULTURAL IMPACTS
Agricultural production, even in the most advanced countries, is
heavily influenced by factors that are beyond the producer's control.
Within the more industrialized countries, yields have increased more
slowly over the past decade than before. This may be partly because of
man-induced environmental factors, possibly including lower air quality,
at least in particular regions. Some efforts have been made in the past
to calculate yield reductions in such regions and then these reductions
have been multiplied by crop prices to estimate the value of lost
production. This apparently straightforward procedure applied in the
past is, however, too simplistic and may very well lead to deceptive
results.
The reason for this is that some particularly high value
agricultural crops, such as vegetables and fruits, tend to be
concentrated in particular geographical regions due to specific climate
requirements. Given the concentration of such production, and the known
adverse effects of air pollution on vegetables and fruits, one might
expect price fluctuations for such commodities in response to changes in
air quality. The same might occur with more generally grown field crops
if the pollution effects are widespread. Any reduction of yields due to
air pollution may affect consumers and producers of those commodities
differently. That is, if the quantity demanded is not very responsive
to price for, say, celery, consumers would suffer a net loss, while
producers in general will benefit from the increase in the price of
celery resulting from the reduction in supply.
This seemingly perverse result invites introduction into the
discussion of another basic idea from demand theory. The relationship
between changes in quantity demanded and changes in price is called by
economists a price elasticity," or "elasticity of demand." Demand
elasticity in quantitative terms is the percentage change in the
quantity demanded divided by the percentage change in price.
Thus, if price goes up by one percent and the quantity goes down by two
percent, the price elasticity is two, and we say that demand is
relatively elastic. If the percentage change in price and the percentage
change in quantity are the same, we say that demand elasticity is
unitary; and if the percentage change in quantity is less than the
percentage change in price, we say that demand is relatively inelastic.
If demand is relatively inelastic, a reduction in quantity will increase
total revenue of producers--the situation cited
30
-------
above. For example, let us say that the price of a commodity is $10
and that at that price the quantity demanded is 20 units. Therefore, in
accordance with the explanation given in chapter 3, the total revenue to
sellers would be $200. Now suppose that the quantity offered for sale
drops by ten percent, that is, to 18 units, but the price rises by
twenty percent, that is, to $12. Then the total revenue would rise to
18 x $12, or $216.
We can illustrate this situation more generally by constructing a
hypothetical demand curve for celery. In this illustration, area P2, D,
Q 2, 0 is larger than area Pi, C, Ql, 0. This means that with
quantity reduction from Q 1 to Q2' total revenue increases.
While producer's profits may grow because of the higher price of
celery, consumers will lose the consumer's surplus shown by the area
(a), (b), (c), and (d) on the diagram. This is the maximum they would be
(d)
P
2
(b)
P
1
0
Q of Celery
willing to pay to avoid the air pollution. This also illustrates the
idea of "derived demand" introduced but not explained in chapter 2.
The willingness to pay for air quality is not, in this case, because
consumers value the air quality directly, but because it is an input to
something they do value--celery.
If demand for celery is relatively elastic, the quantity reduction
would result in both a loss of consumer's surplus and a loss of
producer's profits. In this case, the benefit from reducing air
pollution consists of both the gained consumer's surplus and the
increased profit to producers.
Where price effects of the kind described may be important, it is
necessary to develop a method which can properly handle them in the
process of analyzing economic losses in agriculture from air pollution.
The agricultural component of our research project developed such a
method and
31
-------
applied it to Southern California. This case study is reported in
chapter 9. A different method to handle the same problem for the study
of national economic damages to major field crops, reported in chapter
10, was developed and is explained there.
RESIDENTIAL PROPERTY VALUES--A SUMMARY MEASURE?
In an effort to get a summary measure of the value people place on
cleaner air, economists have developed a method called the "property
value method" for application in urban areas. The general idea is to
assemble information on all the various characteristics which might
determine house price (location, lot size, number of rooms, etc.), on
characteristics of the owner (chiefly income), and on pollution levels
at the site studied. Then, by using the statistical technique
(regression analysis) referred to earlier, and explained briefly in the
next chapter, it is possible to make an estimate of that part of the
difference in house prices which is separately associated with
differences in air quality at the different sites. Through a procedure
which is a bit intricate, and which we need not review here, these
estimates can be used to estimate a summed up (aggregate) "demand" for
air quality in the city or metropolitan area being studied. The word,
demand, is in quotation marks in the previous sentence because economic
theorists have deter-mined that only under a particular set of
circumstances can that number be regarded as a valid and accurate
estimate of the actual willingness to pay for an improvement in air
quality. Nevertheless, the method has some very appealing qualities.
It is relatively inexpensive to do because it can rely on existing
data rather than requiring the collection of new data, which tends to be
quite expensive. That is not to say that existing data are
necessarily high quality, but it can be claimed that the data available
for the case studies using this method, reported in chapter 7, were
quite good.
Also, if such an estimate can be regarded as accurate, it provides
a quick summary measure of the value of air quality to people without
the necessity of estimating the value of different characteristics
individually. These would include effects on visibility, on soiling and
materials damage, and to the extent they are understood, on health.
It would therefore be very useful to run an experiment where
demand estimates derived from property value data are compared with the
actual willingness to pay for improvement in air quality. This
would permit a test of how importantly the theoretical conditions
required to make them precisely equal affect the actual outcomes in
practice. It is, of course, impossible to do this, because there is no
way to get an estimate of willingness to pay which can be regarded as
entirely accurate. This is illustrated by our discussion of possible
biases in bidding games earlier in this chapter.
What is possible, however, is to compare, admittedly imperfect,
estimates made by both techniques for the same people and the same area
to test whether they come out pretty close together or yield
wildly different
32
-------
results. This is the main point of the South Coast and San Francisco
Air Basin studies reported in chapter 7. I turn now to the case
studies.
33
-------
PART II: CASE STUDIES
In the quantitative work done to implement the concepts and
procedures discussed in earlier chapters, a number of case studies were
conducted. The methods and results of these are presented in this part.
The first section in this part deals with urban air pollution and
drinking water considerations, and the second with rural air and water
pollution.
34
-------
Ila. URBAN AIR POLLUTION
The exposure of concentrated populations to air pollution
inevitably raises questions about possible health impacts. Two of the
five cases in this part deal with that issue. Two other cases deal with
the damaging effects of urban air pollution more broadly and are
designed primarily to test the comparability of two quite different
methodologies for assessing benefits from improved urban air quality,
bidding games, and property value studies. The final case in this
section examines the relationship between wage differentials among urban
areas and their levels of air pollution as a possible means for
evaluating air quality deterioration.
35
-------
CHAPTER 5
AGGREGATE EPIDEMIOLOGY--THE SIXTY CITIES
THE SIXTY CITIES STUDY
The first study undertaken by the research team sponsored by EPA
was concerned with the possible linkage between air quality and health.
Its objective was to improve estimates of how air pollution may be
related to increased risk of death (mortality) from various diseases and
to calculate what the economic benefits from reducing this risk might
be. As explained in earlier chapters, doing this requires three main
classes of information: (1) establishing the link between ambient
conditions and effects on humans (the dose-response relationship); (2)
determining the population at risk; and (3) valuing the economic benefit
from improvement in air quality.
The method used to develop the first type of information in this
study is one called "aggregate epidemiology." Ideally, in applying
epidemiological techniques to the air quality problem, one would wish to
have information about the history of exposure of individual persons to
air pollution. Furthermore, to isolate the effect of air pollution from
other factors influencing health, one would wish information about the
individual's personal characteristics, for example, their age, their
access to medical care, health-influencing genetic factors, dietary
habits, whether or not they smoke, and perhaps other pertinent data. At
the time the sixty cities study was done, data including this kind of
information for individual persons were difficult, if not impossible, to
get. Accordingly, resort was often made to other more aggregated, and
therefore less suitable, data.
For example, in the case of the study reported here, information
pertaining to entire cities was used. For instance, total mortality
divided by the population for an entire city was used in the estimation
of a dose-response function. This assumes that the average situation can
represent individual circumstances and responses. This is a
strong assumption which may bias the results of the analysis to an
unknown extent.
Other studies have used aggregate epidemiology to try to understand
the relationship between air pollution and health. The study
described here differed from the others available at the time it was
done in two principal ways. First, it includes factors which are
thought to be health-related, but which others had excluded due to
poor or unavailable data. These are primarily diet, smoking, and the
availability of medical care. Second, in contrast to the usual
epidemiological studies, the present study assumed that people do not
merely accept passively exposure
36
-------
to air pollution, but may take actions to avoid its effects.
In that sense, it is more economic in its orientation in that it
recognizes that there are tradeoffs between air pollution effects and
other values having economic content, for example, incurring the expense
of seeing a doctor or moving away from a polluted city.
It is primarily in respect to such economic-behavioral responses that
economists can make a contribution to the study of epidemiology.
Conventional epidemiology tends to neglect the fact that people have an
incentive to, and do, adapt to environmental conditions.
Instead, it treats them as passive acceptors of whatever occurs.
Before proceeding to further discussion of these matters, it may
be helpful to readers not familiar with regression analysis (mentioned
several times before and a basic tool in this study and many of the
other case studies) to say a little about the technique.
In doing this, and relating the discussion specifically to the present
case, it will be helpful to refer to the figure below.
Along the top, in bold letters, is depicted a
MORTALITY PATE F(MEDICAL CARE, AGE, GENETIC FACTORS, BEHAVIOR & HABITS, DIET, EXPOSURES)
I
Heart Disease
Cancer
Vascular
Disease
Pneumonia
Influenza
Cirrhosis
Emphysema &
Bronchitis
Kidney
Disease
Congenital
Anomalies
Diseases of
Early Infancy
I
Doctors/
Capita
Hospital
Beds/
Capita
I
Median
Age
I I
Rice Smoking
Room Density
Rac,-
I
V I t,-l
Saturated
Fits
Cholesterol
Protein
Additives
Alcohol
coffee
simple equation which says that mortality is related to (or, as
economists say, is a function of) a variety of health-related factors.
In such an equation, mortality and the health- related factors are
called variables. They are called this because the data on them may
take on a range of different values, depending on the particular
situation. For example, where the city is the unit of observation, as
in our study, both mortality and air pollution differ considerably from
city to city. It is these differences that permit regression analysis to
work. It can be viewed as complicated kind of averaging procedure that
uses concepts based on statistical probability theory.
The variable on the left-hand side of the equals sign is called the
dependent variable. This is the variable whose behavior one is trying to
explain. The variables on the right-hand side are called the independent
37
-------
variables, i.e., the ones that are thought to determine what value the
dependent variable takes on.
The goal of the mathematical manipulations involved in regression
analysis is, given the data on hand, to identify and quantify what
separate and independent quantitative effect each of the independent
variables has on the dependent variable. One condition for this process
to work accurately is that the independent variables must not be
interrelated (i.e., themselves correlated). This is almost never the
case with real data, and for this, as well as some other reasons, there
is always more or less uncertainty about the results achieved.
One such instance of interdependency among variables of particular
interest in the present study is the effect of medical care on health.
The existing epidemiological literature has failed to show any
significant effect of medical care on human mortality rates. This
result, which most people would not expect, may have a simple
explanation. For example, in our analysis of sixty cities, no effect of
the availability of doctors on mortality was shown when a
straightforward regression is done where the actually observed number of
doctors in different cities is entered as one of the independent
variables. A possible explanation for this is that, although
availability of doctors most likely does reduce mortality rates (as
shown below), doctors prefer not to live in polluted cities. Therefore,
relative to the total population there are fewer doctors in such cities.
Thus, whatever favorable effects doctors have on mortality rates tend to
be canceled by their fewer numbers in those cities. Simple regression
analyses cannot untangle the relation of doctors to mortality versus the
relation of pollution to mortality. This kind of problem is known
to aficionados of such things as "simultaneous equation bias."
To get to this problem, a "two-stage" regression technique was
used in which one first estimates how many doctors there would be in a
city aside from the influence of pollution, but with other factors the
same as otherwise. Then in the second stage, that estimated number of
doctors is entered into the analysis rather than the actual observed
number of doctors. This technique separates the influence of doctors
from the influence of pollution. This recognition of one aspect of
human adaptation to pollution had a dramatic effect on the results.
For the full-scale analysis, it was possible to develop for a set
of sixty cities the variables shown in the above equation. The dietary
and smoking variables had to be estimated quite crudely, since there
exist no actual observations on them. For example, cigarette consumption
for a particular city was calculated from cigarette sales tax data for
the state in which the city was located. Surely one cannot make any
great claims for the quality of these data. It was felt, however, that
these variables were potentially so important in influencing health and
mortality that to exclude them would be inviting even more serious
error.
This leads to a further observation on regression analysis.
In the language of the trade, a regression equation must be "specified"
properly if one is to have any confidence at all in the result. That
means that the
38
-------
"correct" set of variables must be included in the analysis.
If all the significant variables are not included in the equation, the
equation is misspecified, and a variable that is there may pick up some
of the effect actually attributable to one or more of the missing ones.
For instance, if smoking is importantly related to health, and if,
further, there is a correlation between smoking and air pollution, then
if smoking is excluded, there is, so to speak, a surplus effect to be
picked up and the air pollution variable will take some of it.
Let us then take a deep breath and turn to a discussion of the
results of investigation of air quality dose-response relationships.
The table below is a summary of the signs the various variables took in
regression analysis. In the table, each column represents a regression
equation for a
Variable
(Sign of Hypothe-
tic nl Effect)
Total
Mortal-
ity
Rate
Vas-
cular
Disease
Heart
Di-
sease
Pneumonia
and
Influenza
Emphysema
and Bron-
chitis
C i r r -
hos is
Kidney
Di-
sease
Congenital
Uirtli
Detects
Early
Infant
Di-
seases
Can-
cer
Doctors/^
Capita
(-)
-
-
-
-
-
-
Median Age
(+)
+
+
+
+
+
X Noowhite
(+)
+
+
+
-t-
Cigarettes
(+)
+
+
+
Room Density
(+)
+
+
+
Cold
(+)
+
+
+
Animal Fat
( + >
+
Protein
(*)
+
+
Carbohydrates
(?)
-
NO ^
(+)
(+)
Particulates
(+)
4-
R2
.82
.60
.77
.54
.39
.64
.54
.22
.55
.86
*
Two-stage estimator employed.
cause, or set of causes, of mortality. A
positive (plus) sign means that an increase in the level of the variable
tends to increase mortality, and a negative sign (minus) means that an
increase in the level of the variable tends to reduce mortality.
The results shown here are only the "significant" ones.
That means that they have passed a purely formal statistical test, but
in view of the difficulties of equation specification, simultaneous
equation bias, and other problems in application, it does not
necessarily mean that they are "true." But if one is confident that the
equation is specified about right, and the results for a particular
variable are fairly large (and "statistically significant"), it means
that the hypothesis that the relationship is real and about of the
magnitude estimated cannot be ruled out. Associated with each
variable in the regression is a number called a
39
-------
regression coefficient. This number is used to quantitatively estimate
the change in the dependent variable when the level of the independent
variable changes, for example, when air pollution goes down. Since they
are not used explicitly in this discussion, these numbers are not
reported here.
Let us look more closely at the results. Both the median age and
percent nonwhite variables are widely significant across the estimated
variables, and show up with uniformly positive effects on mortality
rates.
Cigarette consumption shows significant positive relationships
with total mortality, vascular disease, heart disease, and cancer, while
room density (average number of persons per room) and cold (number of
days in which temperature drops below a specified level) both show
significant positive relationships with total mortality, and pneumonia
and influenza. Room density also shows significant positive
relationships for cirrhosis and kidney disease.
The dietary variables show significance in total mortality, heart
disease, and cancer--relationships between heart disease and saturated
fats and between cancer and meat consumption (note the positive
association for protein) have long been recognized. The dietary
variables also show up as significant in emphysema and bronchitis. Not
much credence should be given to the individual dietary variables,
because the data are poor and the variables are highly interrelated.
Our main concern with diet in this analysis is that we have accounted
for diet in a general way in specifying an equation where the primary
interest is in the air pollution variables.
Turning to the air quality variables, only two significant
correlations appear--between particulates and the pneumonia and
influenza variable, and between sulfur dioxide and the early infant
disease variable. It should be observed that these associations we have
found between mortality and air pollution are primarily for diseases of
the very young and very old--particularly susceptible groups within the
population. Further, these effects are those which one would usually
associate with short-term as opposed to long-term air pollution
exposures. We have some confidence in these particular results.
It may well be that aggregate epidemiology may be incapable of
revealing the long-term consequences of air pollution exposures, if they
exist. A reason could be that data on the actual air pollution exposure
history of people are not available. In view of both changes in
environmental conditions and the mobility of the population, current
observations of ambient air quality may simply not be an adequate
indicator of actual exposure to capture any effects of air pollution on
degenerative diseases. For example, cancer may occur as long as two
decades after exposure to carcinogenic substances.
Having investigated the dose-response relationships, we now turn
to an economic evaluation of air quality control as it pertains to
reduced mortality. This analysis is based on the valuation of risks
approach discussed in the previous chapter. As the reader will recall,
figures were quoted there which various scholars had obtained by
analysis of risky
40
-------
occupations. Two of the estimates are presented in the following table,
along with a sketch of the methods described below.
Benefits = (Population at Risk) x (Value of Safety) x
(Reduction in Health Risk)
Value of Safety Based on a Consumer's Willingness to Pay
Low estimates: $340,000
Source: Thaler & Rosen (1975)
High estimates: $1,000,000
Source: Robert Smith (1977)
(Other recent estimates have even higher range.)
First, to obtain national estimates, we must know, as explained in
the text in chapter 3, the population at risk. Since our sixty city
sample is entirely urban, and since air pollution-related health effects
is principally an urban problem, we used a population at health-related
risk of 150 million urban dwellers. As a range for the value of reduced
risk, we used Thaler and Rosen's (1975) estimate of $340,000 per life
saved as a lower bound, and Smith's (1977) estimate of $1,000,000 as an
upper bound. Finally, to get an estimate of reduced risk from air
pollution control, we assumed an average 60 percent reduction in ambient
urban concentrations both for sulfur oxides and particulates. Then,
using the average concentration of these pollutants in our sixty city
sample as a basis for calculation, we derived an average reduction in
risk of pneumonia mortality for a 60% reduction in pollution from our
estimated dose-response functions for these diseases. It should be
noted that this is a very large reduction from present levels, and it
would be difficult and very expensive to achieve.
Note, in terms of our discussion of chapter 2, that a more
complete analysis would have assumed various levels of control at
different emissions points and then used a dispersion model, described
there, to calculate changes in the population at risk and levels of
risk. We would have then used these, along with the dose-response
estimates and risk valuations, to calculate associated benefits. The
capability of doing this on a national scale does not yet exist. It
would be a monumental job to achieve a high level of accuracy in such an
undertaking for the entire nation. But this type of capability does
exist in various regions.
Multiplying the population at risk by the assumed value of reduced
risk, and then by the average reduction in risk, gives a crude
approximation of the benefits for a 60% reduction in national urban
ambient concentrations of particulates and sulfur oxides, respectively.
National
41
-------
urban totals and the value of the average individual risk reduction are
displayed in the following table.
Disease
Pollutants
Average Individual
Safety Benefit
(1978 $/Year)
National
Urban Benefits
(1978 $billion/Year)
Pneumonia
Particulates
29 - 92
4.4 - 13 . 7
Early Infant
so 2
5 - 14
.7 - 2.2
Disease
TOTAL
34 - 106
5.1 - 15 . 9
If these results are accurate, and we cannot vouch for how
accurate they may be, they are, in our judgment, conservative.
This is because we believe that air pollution may well have long-term
chronic health effects, and some evidence to that effect is established
in the next chapter, but that, given the available data, aggregate
epidemiology cannot dependably establish them. Moreover, while
this study explicitly recognized the specification problem, and made
some progress in dealing with it, it did not address the problem in a
fully systematic manner. This creates further uncertainty about the
accuracy of results. In the next chapter, I turn to a very recent study
that promised to be capable of yielding much more reliable results.
42
-------
CHAPTER 6
DISAGGREGATE EPIDEMIOLOGY AND MORBIDITY
FROM OZONE EXPOSURE
The most general conclusion to be drawn from the experiment with
aggregate epidemiology discussed in the previous chapter, and from
review of the work of others, is that the results of such studies, if
accurate at all, are so only within very wide bounds. It appears that
only observations on individual persons might improve the situation.
We refer to epidemiology using econometric techniques such as regression
analysis, and performed on data about individual persons, as
microepidemiology. The objective of the microepidemiology study
reported here was to examine a possible link between exposure to ozone
and health damage. The known physiological effects of ozone suggest that
such a link could exist.
For this purpose, a large amount of effort was expended to build a
suitable data base (data base is a general term that usually is used to
refer to an assembly of quantitative information suitable for analysis).
This was accomplished by merging the information in two existing sets of
data that have recently become available. The first is the 1979 Health
Interview Survey (HIS) data assembled by the National Center for Health
Statistics. The second is 1979 air pollution data from EPA's monitoring
system. The resulting data base contains a very large number of
observations, a situation which, as we will see shortly, has a number of
advantages.
The HIS was started in 1957 and has been conducted annually since
then. A sample of about 110,000 people located across the country is
interviewed every year. A number of types of information pertinent to
epidemiological study are obtained. Among them:
(1) demographic characteristics (including age, income, sex,
occupation, etc.)
(2) number of days during the two-week period prior to the survey
on which the respondent had to restrict his or her activity,
stay in bed and miss work or school
(3) visits to doctors or dentists during this two-week period
(4) acute and chronic health conditions (including some
diagnostic information) accounting for restricted activity or
doctor visits
43
-------
(5) hospital episodes during tile twelve-month period prior to
the interview
(6) smoking habits, history of residential mobility (the
preceding two items were elicited only in the 197 9 survey) ,
home health care utilization, vaccination history, eye care,
and retirement income.
As far as air pollution data are concerned, information was
collected from EPA's System on ambient concentration for eight major
pollutants for 1979. Using a program that matches individual census
tracts to the nearest air pollution monitor, it was Possible to assign
each individual in the HIS the air pollution readings nearest to his or
her home. While this still does not provide an accurate measure of an
individual's dose, it is clearly superior to the assumption that an
entire city is subject to the same exposure, a device that had to be
resorted to in the macroepidemiological study reported on in the
previous chapter. Also, the data set permitted matching individual
health status to air quality conditions prevailing for two or three
weeks prior to the interview as well as to annual averages and other
periods. This is important in trying to identify acute effects.
The final set of data contained about 14,500 adults and about
15,7 00 children. These were persons who could be matched with air
quality monitors and for whom smoking information could be obtained.
In the previous chapter, it was pointed out that one of the factors that
limits macroepidemiological analysis is the mobility of the population
which means that exposure at their current address often reflects long-
term exposure only very poorly. A supplement to the 1979 HIS interviews
provides information on persons who have resided at the same location
for a long period of time. The large size of the whole sample permits
these persons, whose exposure history can be defined more accurately, to
be studied separately.
An unfortunate aspect of the data set is that it does not permit
study of the possible linkage between air pollution and mortality, the
relationship analyzed in the macroepidemiological studies. Therefore,
direct comparisons of the two approaches are not possible. But the data
set does contain information about the presence or absence of chronic
respiratory, cardiovascular, and other illnesses. Thus, it was possible
to test the proposition that air pollutants not only may induce acute
health effects, but may be related to the development or prolongment of
chronic conditions which may lead to earlier mortality.
Another advantage of the very large data set is that effects on
children, asthmatics, or other potentially sensitive people can be
tested. Other efforts to do this have been hampered by small sample
size. Many such subsamples were analyzed to test the sensitivity of
results to the kind of group selected. For example, some public health
specialists believe that environmental pollutants can have an especially
adverse effect on children. Experiments to test this hypothesis were
done, and they did not confirm it in the case of ozone. Accordingly,
the remainder of the
44
-------
discussion in this chapter will focus on the work done trying to
identify impacts on the entire set of adults.
In accordance with information available from the HIS surveys, the
researchers tried to identify relationships between ozone levels and
four main dependent variables (recall the discussion of dependent and
independent variables in the previous chapter) . Three had to do with
acute illness: these are "restricted activity days," "work loss days,"
"bed disability days." Finally, the researchers examined the information
for a link between ozone and "chronic respiratory disease." The
independent variables in general resembled those used in the "sixty
cities" study discussed in the last chapter, for example, age, sex,
income, and smoking habits, among others, except that in this instance,
they pertain to actual individual persons in the sample rather than
being averages across entire cities. Having information about individual
persons also permitted introducing some other variables which could
influence morbidity. One of these, labeled "FAT," was meant to represent
the person's general physical condition. It was possible only to
construct a crude proxy, and this consisted of weight in pounds divided
by height in inches. The hypothesis was that being excessively
underweight or overweight might be associated with higher morbidity. In
general, results of the analysis supported this view.
One independent variable that deserves special note in the
analysis of acute morbidity is "chronic illness." Here the hypothesis is
that persons who have a chronic disease may be more subject to episodes
of acute morbidity than those who do not. Again, the analysis conducted
is consistent with this supposition. Of course, in each instance, ozone
levels at the appropriate monitoring stations for each individual were
included among the independent variables, as were readings for the other
air pollutants.
The method used to try to identify the separate effect of each of
these independent variables on the dependent morbidity variables was
once again the regression analysis explained in the last chapter. A
number of experiments were conducted with regression equations,
including different variables (specifications) and different subsamples
of the whole sample.
As one would expect from the discussion of the last chapter, the
results were sensitive to the specification used. For example, the
calculated effect of ozone on health was often influenced by what other
pollution variables were included in the equation, for example,
particulate matter.
But for the acute morbidity variables--restricted activity days,
work loss days, and bed disability days--the relation between them and
ozone readings was almost always positive, and in many cases,
significant, in the purely formal statistical sense noted in the last
chapter. People in the econometrics trade refer to a variable that
behaves in this consistent manner as across experiments being "robust."
Robustness in the uncertain world of statistical analysis gives one some
confidence that what is being observed in the data is real. Although
they find some very tentative
45
-------
evidence to support it, the relationship between ozone levels and
chronic respiratory disease is less robust and therefore leaves one in
greater doubt as to its genuine existence.
Thus, even epidemiology conducted with information on individual
persons and with much better exposure data than is available for
macroepidemiology does not yield the clean and persuasive results one
would wish and hope for.
With these cautions in mind, let us turn to some sample results
concerning the changes in health status that a change in ozone levels
might yield. Once a relationship between independent and dependent
variables has been estimated, it is then possible to hypothetically
change the value of an independent variable and, using this quantitative
relationship, calculate the associated change in the dependent variable.
For example, one can reduce the observed ozone level and calculate the
effect on, say, restricted activity days.
I will present the results from two of the many equations
estimated. These may be taken to be representative of those in which
the association between ozone and health indicators was positive and
significant in the statistical sense.
The first illustration is an equation which tries to establish an
association between ozone and restricted activity days (RAD).
It is specified as follows:
RAD = F(ozone, sulfates, race, sex, marriage status, income, urban, FAT,
age, smoking, education, chronic health condition, crowding,
temperature, precipitation, humidity)
In this equation, ozone and several other independent variables,
for example, income and chronic disease, meet the statistical
significance test, but most of the others do not. As an
unexplained anomaly, the sulfate variable is negative and significant.
This has the hardly credible implication that sulfur pollution is good
for you.
If one takes the ozone result at face value, it indicates that
each .01 part per million increase in the highest hourly reading for a
day could result in 0.25 more RADs per person over a two-week period.
(Recall that respondents to the HIS survey were asked to report on
their health status for the two weeks immediately prior to the survey.)
Or conversely, a similar decrease in ozone would result in a similar
decrease in RADS. If this result is extrapolated to a whole
year, it means that .01 PPM reduction in ozone would result in
.64 fewer RADS, on the average, per person per year.
Extrapolating further to a population of about 110 million adults
(over 17) in the metropolitan areas of the United States
46
-------
implies about 7 0 million fewer RADs for the country, associated with the
ozone decrease. The study reported here made no effort to assign an
economic value to this environmental improvement, but it is immediately
apparent that the per day value would not have to be large to yield an
impressive annual benefit. If , for example, the consumers surplus from
avoiding a RAD were a mere $10 per day, the aggregate national benefit
would be about 700 million dollars per year.
The other illustrative results pertain to the possible effect of
ozone on chronic respiratory disease (CRD). The equation specification
for this analysis was:
CRD = F(ozone, suspended particulates, sulfur oxide, race,
sex, marriage status, income, FAT, age, smoking,
education, temperature, precipitation, humidity)
In this case in addition to ozone, sulfur oxide (this time
positive, but small), race, sex, income, age, education, and humidity
were significant in the statistical sense.
If one goes through calculations in spirit like those outlined for
RAD above, the results show that .01 PPM reduction in the average annual
hourly ozone concentration across the country would cause the incidence
of CRD to gradually evolve to a point where there would be about
1,130,000 fewer cases per year. If one assumes that the consumers
surplus associated with avoiding one case of CRD is only a thousand
dollars, the .01 PPM reduction would ultimately yield benefits of
greater than one billion dollars a year. The full benefits are not
available immediately because chronic disease lags exposure change by
some years.
There are several things to be said about these results: first,
while .01 PPM is, from an everyday perspective, a very small number, it
is large compared to the actually existing average levels of ozone.
It implies about a 20 percent decrease from the average daily maximum
reading around the country and a 50 percent decrease from the average
hourly reading. It should be noted that such a decrease would be
difficult and costly to achieve. It is not necessarily obvious that
even if the benefit numbers quoted above were true the economic benefits
would outweigh the costs.
Second, as noted, the results are rather robust with respect to
the statistical linkage between ozone and acute morbidity. It does
not seem unreasonable to argue that it provides support for the view
that there is a significant real link. The results with respect to
chronic respiratory disease are not robust and therefore have to be
doubted.
Third, in general the results of the macro- and microepidemiology
studies reported here, and related work of others, do not foreclose the
possibility that health benefits from reducing air pollution are large.
47
-------
But the exact, or even approximate, magnitude of those benefits is far
from being established.
48
-------
CHAPTER 7
AIR QUALITY BENEFITS IN THE SOUTH COAST AIR BASIN
AND IN SAN FRANCISCO
THE BASIC STUDIES
For the household sector, and considering other factors in
addition to health, two distinct approaches to valuation of
environmental quality have emerged from recent research. The first, as
explained in chapter 4, involves the analysis of how some pertinent
actual market prices, such as real property prices, are influenced by
environmental quality attributes of the properties. The second, also
discussed in chapter 4, tries to induce individuals to reveal directly
their actual preferences in monetary terms for environmental attributes.
Clearly, if these methods are valid, there should be a well-defined
relationship between what people do pay through differences in property
values and what they say they will pay, provided there are no incentives
for them to distort their bids and that influences other than air
quality on property values are correctly accounted for.
The first study area where these techniques were tested and
compared in our study--the South Coast Air Basin (SCAB)--consists of
Orange and Los Angeles Counties and portions of San Bernadino and
Riverside Counties of California. This area has a long history of air
quality problems. For instance, Spanish explorers in the sixteenth
century noted smoke from Indian campfires in the Basin, trapped by
inadequate horizontal and vertical air mixing. The post-World War II
period, which saw extremely rapid population growth in Southern
California accompanied by massive industrial development, was marked by
the appearance of smog as the major threat to the regional environment.
As a result, air pollution abatement programs began in the late 1960s as
a response to the discovery of the automobile's role in smog formation.
Air quality deterioration in the SCAB has multiple causes: unfavorable
topography and meteorology, and dense population and economic activity
with corresponding large emissions.
To conduct the study, a special sampling procedure was developed.
It was designed to identify paired communities in the Basin that are
similar in as many ways possible except in air quality. If the other
characteristics of these communities are not very different across areas
(housing styles, sizes, distance to the beach, etc.), the difference in
property values between an area characterized by clean air versus an
area where air quality is lower should be due mostly to the existence or
absence of pollution. This structured, paired communities sampling
procedure, rather than a random sample of individuals over the whole
region, was chosen primarily to control for proximity to the beach.
Nearness to the ocean and
49
-------
cleaner air are so highly correlated that the most applicable
statistical procedure, regression analysis, performed on a random sample
would not be able to disentangle these two major influences on house
prices in Southern California.
The Los Angeles area was chosen for the initial experiment not
only because of the well-defined air pollution problem there, but also
because of the existence there of excellent property value data.
Twelve census tracts were chosen for sampling for both the property
value and 'the companion bidding game study. For the latter, interviews
were conducted in these tracts during march 1978. Respondents were
asked to state their willingness to pay for an improvement in air
quality at their current location. Air quality was defined as poor,
fair, or good, based both on maps of the region (the pollution gradient
across the area is both well-defined and well-understood by local
residents) and on photographs of a distant Vista representative of the
differing air quality levels. Households in poor air quality areas were
asked to value an improvement to fair air quality, while those in fair
areas were asked to value an improvement to good air quality. A total
of 290 completed surveys was obtained. The map below shows the areas
having poor, fair, and good quality air in the South Coast Air Basin.
Cjho ;» ?iric
r-
PO 0 R
G00.C
FAIR
CCt
GOOD
• iMiMtit tie w«Mr«rtJtc
Built into the survey questionnaire were procedures for
identifying the various possible biases in bidding games, as discussed
in chapter 4. No
50
-------
biases were found. The results indicate that on the average, households
said they were willing to pay $30 per month more for the cleaner air
areas. For comparison to the survey responses, data were obtained on
634 single family home sales which occurred between January 1977 and
March 1978 in the paired communities used for the survey analysis.
Households will choose to locate somewhere along a pollution-property
value gradient paying more, other attributes being equal, for homes in
clean air areas, depending on their family income and tastes. However,
economic reasoning suggests that cost difference between homes in two
different air quality areas will exceed the willingness to pay as
elicited by a bidding game for similar improvement in air quality. Thus,
we would expect house cost difference associated with air quality
improvement to exceed estimates of household willingness to pay from the
survey responses. This is because property values at a particular
location will reflect the air quality preferences of the most air
quality- sensitive individuals, whereas average bids for that same air
quality will more nearly reflect the average preferences of people
living there. Most houses are not for sale at a given time, but given
the small number available, their price will be determined by those who
want them most, for example, those people with the strongest preference
for cleaner air.
A straightforward statistical comparison of the paired
neighborhoods indicates that property value differences between poor and
fair air quality localities are about $140 per household when computed
on a monthly basis. Using more advanced economic models, which better
take into account factors other than pollution, such as any remaining
influence of distance to ocean and differences in tastes which may
influence property values, willingness to pay inferred from the property
value differences is about $4 0 per month. As a reasonably comparable
estimate, the survey results, as indicated, show an average bid of
slightly less than $3 0 per month.
The results indicate that air quality deterioration in the Los
Angeles area has had substantial effects on housing prices and that
these are comparable to what people say they are willing to pay for
improved air quality. Moreover, the property value estimates are higher
than the average bids, which, as noted above, was expected on
theoretical grounds.
Based on these results, rough estimates can be made about
willingness to pay for improved air quality throughout the South Coast
Air Basin. Difficulties are encountered in making data sets for groups
of diverse households exactly comparable. Significant differences exist
between the people in the survey and the property value groups in
average income, age, and other socioeconomic factors. Accordingly,
any extrapolation to the Basin as a whole must be taken as rather crude
and merely indicative rather than exact.
The following table gives estimates of monthly bids for cleaner
air by households, results of the property value study, and by
extrapolation of the benefits for an approximate 3 0 percent improvement
in air quality within the South Coast Air Basin. The latter estimate,
while quite rough, does suggest that economic benefits from an
improvement in air quality in the South Coast Air Basin are very large.
51
-------
(3 0 percent improvement
in air quality)
Annual benefits (in
billions of dollars)
1977 Dollars
for the South Coast
Air Basin
Property Value Study
Based on straight- $135 $3.96
forward comparison
of communities
Calculated $ 42 $ .95
willingness to pay
taking account of
other factors
Survey Study
Mean Bid $ 29 $ .65
The results of this experiment also suggest that survey
instruments, when compared with property value techniques, may provide a
reasonable way to get environmental quality benefit estimates. The
survey approach has the advantage that new data can be collected at low
cost on specific environmental problems. The investigator is not tied
to the availability of existing data sets which are usually not designed
to meet his particular needs.
As a caution, however, it should be kept in mind that the South
Coast Air Basin studies were conducted in an area where individuals have
both an exceptionally clear-cut pollution situation that they have
themselves experienced and where there exists a well-developed property
value market for clean air. The effect of clean air on property values,
and in turn, on the degree to which people are aware of increased
housing prices in high air quality areas, appears to be exceptionally
well-defined in the South Coast Air Basin. Therefore, it should be
recognized that the results of this experiment may well not carry over
completely to other situations where air quality is not so well-
specified, either through actual market prices or the perceptions of
people.
In view of the possible uniqueness of the Los Angeles Basin, a
follow up study was done replicating as much as practical the Los
Angeles study. The place chosen was the San Francisco Bay area.
This is a large shallow basin ringed by hills stretching from Southern
Marin to Santa Clara Counties. The basin tapers into a series of
sheltered valleys including Santa Clara, Livermore, and Napa.
While the area typically has better
52
-------
ventilation than the Los Angeles Basin, still this topography gives the
area great potential for trapping and accumulating air pollutants.
The map below shows the study area. The numbers on the lines
indicate increasingly higher levels of smog pollution.
1
iTEO
IANTA
CXABA
20 MILES
33 KM
As in the case of Los Angeles, both property values as they related to
were applied. These were compared with each other, given the hypotheses
explained earlier, and to the Los Angeles results. Let us turn first to
the property value study.
While the intention was to make the two studies as comparable as
possible, there were some inevitable differences in both the situations
and in the data available that made some adaptations necessary. For
example,
53
-------
in Southern California, as suggested, the mild year-round climate
encourages a variety of ocean-related recreational activities. Beach
front activity is highly valued, and beach front property has generally
been densely developed. In the San Francisco area, the Bay is the most
accessible body of water to major population centers; however, the Bay
does not offer the same scenic or recreational experiences found along
the coast of the Los Angeles area. In the Bay area, ocean front property
is located over the ridge of the Santa Cruz Mountains and is less
accessible to the major employment centers. As a result, much of
the beach front property maintains a rural atmosphere.
Accordingly, it was not necessary to adopt the paired communities
approach of the Los Angeles study to control for access to the beach.
This made a more nearly random sampling approach possible, which has the
advantage of providing a more dependable basis for extrapolating the
sample results to the entire area.
Another principal difference between the areas is air quality.
Smog is considered to be the major problem in both regions. The city of
San Francisco itself has a less severe air pollution problem than Los
Angeles. However, some cities included in the region (San Jose and Los
Gatos, for example) suffer from severe pollution problems.
Thus, while the San Francisco region provides suitable contrast
in air quality from place to place, still, air quality degradation is
not in general so severe, and one would expect also possibly not so
well-defined in people's minds. Accordingly, it was judged to be an
excellent place to see whether the Los Angeles results would hold up in
this different situation.
Data on property values were gathered or constructed for 2,500
households in the region. These same households also were subsequently
used for interviewing. In addition, data were collected on about 5,000
residential property sales in areas where these families live.
Unfortunately, the sales data available for San Francisco were not as
accurate as the Los Angeles data. The data were used in regression
analyses to try to isolate the effects of degraded air from other
factors affecting air quality, such as income of residents, house
characteristics, access to work places, etc.
Results of the San Francisco study as compared with the South
Coast Air Basin study were in accordance with the hypotheses made about
them.
First, one would expect a thirty percent improvement in air
quality to yield less benefit per capita in an area where air is already
relatively clean. Two types of results from the San Francisco study
support this supposition: (1) a thirty percent improvement in air
quality yields a much larger benefit estimate, by both the property
value and survey method, for the dirtiest subarea in the San Francisco
region than the average for the region as a whole; and (2) a thirty
percent improvement in air quality, again as estimated by both
techniques, yields an average benefit estimate
54
-------
between five and six times as high in the Los Angeles region as it does
in the San Francisco region.
Second, one would anticipate more variability of results from
subarea to subarea in a region (San Francisco) where the pollution
problem is both less intense and less well-defined than in the Los
Angeles area. Again, this supposition is borne out.
As before, on theoretical grounds, one would suppose that the
property value study would yield higher estimates of benefits than the
survey approach. As in the South Coast Air Basin case, this expectation
is met in the San Francisco study, and the relationship between the two
alternative estimates is about the same as in the former.
Accordingly, the two studies have a broad consistency in that the
differences in their findings are expected differences. In general, the
San Francisco study supports the conclusion of the Los Angeles study
that survey instruments may provide a reasonable, low cost way to get
environmental quality benefit estimates.
AN ILLUSTRATIVE BENEFIT-COST ANALYSIS
While the basic purpose of the studies reported on in this volume
was to develop improved methods for evaluating the benefits from air
quality improvement or maintenance, an illustrative benefit-cost
analysis was also done. This is included simply to illustrate how
benefits estimation fits into the economic analysis of environmental
policy. Not much credence should be accorded the actual numbers.
The subject of the study was the benefits and costs of meeting national
ambient standards in the South Coast Air Basin. Selection of this area,
aside from its intrinsic importance, permitted use of the information
developed for the benefits study reported earlier in this chapter.
The national ambient standards for oxidants (.12 ppm maximum
hourly concentration--since the study was completed, the basis for the
standard has been changed from oxidants to ozone) and nitrogen dioxide
(.05 PPm annual average concentration) are consistently violated
throughout the Basin with the notable exception of the immediate coastal
areas which were characterized in the previous discussion of the Los
Angeles area study as having "good" air quality. Accordingly, if
the entire South Coast Air Basin were to be brought into compliance with
ambient standards, areas that were in the earlier study characterized as
having "fair" or "poor" air quality would then be characterized as
having "good" air quality. The development of an aggregate benefit
measure for achieving ambient standards (note that this is a different
objective from the "thirty percent improvement" assumed in the earlier
study) for the entire basin is then done by extrapolation. Benefits are
taken to be the aggregate willingness to pay for all households in both
"poor" and "fair" air quality areas to have "good" air quality, as
defined for both the property value and survey studies.
55
-------
In making the necessary benefits estimates, the property value
results were the ones actually used. These results allow calculation
of household willingness to pay as related to income and air pollution.
It is this relationship that was used for benefit calculations. It
assumes that income and population affect willingness to pay for air
quality improvement in the same way throughout the Basin as they did in
the limited sample. The estimates are strictly for household
willingness to pay and exclude any agricultural and ecosystem effects.
Agricultural benefits for the area are discussed in chapter 8.
Since benefits were calculated for moving from the current (1976
emissions inventory) level of air quality to the ambient standards,
costs must be calculated on the same basis. However, analysis
indicated that costs for on-road mobile source control measures were
substantially more better done than those associated with stationary
source controls. Therefore, only the costs attributable to on-road
mobile source control were examined in the study. The benefits that are
counted are also, necessarily, then only those corresponding to the
share of total emissions reductions which are accomplished by mobile
source control.
Although a careful engineering cost study of using mobile source
control to achieve ambient standards would have been desirable, the
objective of the study was, as noted, mostly illustrative, and resources
for it were quite limited. The study was therefore forced to use cost
estimates found in literature. Unfortunately, in many cases these are
quite uncertain. For the most part, manufacturers statements and
government publications were relied upon for cost calculations.
In addition, the state of California's Air Quality Management Plan
(January 197 9) was the basis for the calculation of required emissions
reductions--the necessary "link" between emissions and ambient
conditions discussed in chapter 3. Calculations presented in the plan
indicate that to achieve ambient standards in 1979 would require
reductions of about 975 tons per day in reactive hydrocarbons, about
6,000 tons per day of carbon monoxide, and about 500 tons per day of
nitrogen oxides. Of these amounts, it was estimated that mobile source
controls are responsible for about 730 tons per day, all of the
reduction, and about 400 tons per day of hydrocarbons, carbon monoxide,
and oxides of nitrogen, respectively.
Applying these methods and data, it was found that benefits of
achieving ambient standards for air quality in the South Coast Air Basin
for 1979 (note this is a much larger improvement than the 30% reported
in the table above) fall in a range of 1.5 to 3.0 billion dollars per
year. Of this total, on-road mobile source control would be responsible
for approximately 1.4 to 2.6 billion dollars. The corresponding total
basinwide control costs fall in the range of .6 to 1.32 billion dollars.
It therefore appears, with due regard to all of the many uncertainties
involved, that the benefits of achieving mobile sources controls in the
South Coast Air Basin could outweigh the costs.
56
-------
CHAPTER 8
AIR QUALITY, WAGES, AND NATIONAL BENEFITS
FROM URBAN AIR POLLUTION CONTROL
As indicated in previous chapters, one of the lines of study in
searching for improved methods to estimate benefits of air quality is to
look for actual human responses, reflected in prices of things, that
might give a clue as to how much people value clean air.
Application of the property value approach discussed in chapter 6 is one
such effort. This approach is based on the idea that people's
residential locational choices reflect the ambient air quality as well
as a number of other characteristics of particular sites.
Another way in which human behavior, reflected in a price, might
display preferences with respect to air pollution is the differences in
compensation that people might demand for performing particular
jobs at different locations with differing air pollution
characteristics. The idea is that, in considering job and location
choices, workers will take into account pollution in the area as well as
other work place characteristics. One of the studies in the program of
work being discussed here was designed to test this idea.
As is, unfortunately, always the case in this game, the data
available for doing the analysis are far from ideal. A basic source of
information used was the Panel Study of Income Dynamics, sponsored by
the Survey Research Center at the University of Michigan. This study
yielded usable wage information on about 14 00 heads of households across
the country. The information obtained in this survey included the
household's state and county of resident and type of employment.
The location information permitted matching of other information about
variables that might influence real (price-correlated) wages, one of
which might be pollution. As in the case of the epidemiology study
described in chapter 5, a set of "independent" variables was specified
that was thought to influence wages, data about them were developed, and
the regression technique used to try to estimate the separate influence
of each of them on wages.
The general form of the equation used was as follows:
Wage = f(whether the individual is a union member, whether the
individual is a veteran, the size of the individual's family, the
individual's health status, the individual's education, the length of
time the individual has spent an his present job, the climate in
the individual's area of residence, job hazards, and levels of
pollution-sulfur dioxide, total suspended particulates, and nitrogen
dioxide).
57
-------
Of all of the independent variables, the worst quality of data is
for the air pollution variables. This is both because some of the
measurement procedures are not very dependable and because, in some
areas, there are not many monitoring stations, and there may not be one
close to the person's place of work--the pollution data are available
only on a county basis.
Further, uncertainty about the accuracy of results comes from the
fact that, because of data limitation, it was not possible to include
all variables that might influence wages. For example, the availability
of recreational opportunities and social services might influence wage
rates. If variables are excluded that have an important influence, we
know that the results may be biased. In chapter 5, the specification
problem was discussed more extensively in connection with its role in
epidemiology analysis.
These qualifications having been made, the results of the analysis
show that only total suspended particulates are statistically
significantly related to wages. The estimate of this relationship and
of the other variables that influence wages can be used to make an
estimate of the damage avoided (benefit) of reducing suspended
particulates in particular urban areas. This is done by putting the
actual observed value of all the other independent variables for that
metropolitan area in the equation, except that the secondary standard
for particulates is substituted for the actual value, and calculating
the implications for wages using the regression relationships computed
from national data. This result is then adjusted for the size of
population of the particular metropolitan area.
This kind of calculation was done for the Denver metropolitan area
and the Cleveland metropolitan area. The resulting total benefits per
year for Denver were about $240,000,000, and for Cleveland about
$70,000,000.
An attractive feature of the methodology just described is that it
is fairly straightforwardly adaptable to producing a national estimate
of the benefits of pollution control. It is a relatively simple matter
to make an estimate of the type described above for each metropolitan
area in the United States and then to add them up to form a national
benefits total. But doing so would have required more data collection
and calculation than available resources permitted.
A very rough approximation of this procedure can, however, be done
very simply as follows:
3 . These results are roughly consistent with those found in a
related study by another member of the research team--Maureen Cropper.
They are reported in "Inter-City Wage Differentials and the Value of Air
Quality," Journal of Urban Economics, September 1980.
58
-------
First one assumes that the situation in Denver is characteristic
of metropolitan areas in the West, and Cleveland of conditions in the
East. Then one computes the per capita benefits for each metropolitan
area and multiplies the result times the total population of the Western
and Eastern metropolitan areas, respectively. When these calculations
are done, an estimate of yearly benefits of meeting secondary standards
of about $5 billion dollars is obtained for the West, and about $4
billion dollars for the East, and about $9 billion for meeting the
secondary standard for suspended particulates everywhere. This is, of
course, an exceedingly crude procedure, and the amounts given are simply
meant to be illustrative of the method.
We presume that if these figures have any validity at all that
what is being measured is primarily the more visible and tangible
aspects of air quality--visibility and soiling--rather than health
effects. If this is so, the benefits to health from a large improvement
in air quality should be added to these estimates. In chapter 5, we
estimated that such benefits could range between about 5 and 16 billion
dollars per year. If this range is also accepted, our total estimated
urban benefit from a large improvement in air quality might be between
15 and 30 billion dollars per year.
While the basis for these figures is scandalously weak, and they
cannot be put forward as genuine estimates but only as illustrative of
methods, I do not necessarily find them incredible. For example, in the
relatively more carefully done studies in the South Coast Air Basin
discussed in chapter 6, annual benefits from a large improvement in air
quality were, as estimated for the benefit-cost study, in the range of
1.5 to 3.0 billion dollars. If we compare this to the higher of the two
national estimates, it does not seem unlikely that benefits in Los
Angeles could be five to ten percent of the total. While metropolitan
Los Angeles has about 2% of the U.S. population, it also has the
nation's most severe widespread air pollution problem. It does seem
unlikely, however, that the Los Angeles area could have as much as ten
to twenty percent of the benefit from a large improvement in national
air quality to meet ambient air quality standards and protect health.
It therefore seems unlikely, based on this slender bit of evidence, that
the number given is an overestimate of national urban air quality
benefits.
It should be noted in closing that some, possibly important,
benefits are not captured, or not fully captured, by the methods and
data presently available. An example is materials damage which could be
quite large.
Finally, in closing this section on urban damage, the reader
should be reminded that the central objective of the research reported
in this volume is to improve the methods rather than to make actual
estimates. Any numbers presented as illustrious in the text must be
appropriately discounted in light of that fact.
59
-------
lib. RURAL AND REGIONAL AIR AND WATER POLLUTION
The urban cases reviewed in part I la are all instances where one
starts with a degraded condition and wishes to know the benefits of
improving it. The first four cases reviewed in this part are also of
that type. They concern Southern California agriculture, national
agriculture, freshwater fishing benefits from water quality improvement,
and national benefits from water quality improvement. But there are
also very important rural air pollution issues that raise the question
of what it is worth to protect an area that is still relatively
pristine. The cases of this type that are reviewed in the later
chapters of this section are concerned with the matter of protecting
visibility in National Parks, protecting groundwater from
contamination, and with protecting water courses and other parts of the
ecosystem against acid rain.
60
-------
CHAPTER 9
AIR QUALITY BENEFITS TO SOUTHERN CALIFORNIA AGRICULTURE
As indicated in chapter 3, agricultural production is affected by
many influences beyond the control of individual producers. In
agricultural regions within or nearby urban areas, air pollution has, in
recent decades, become one of these influences. As further pointed out
in chapter 4, when these agricultural regions, say because of unique
climate characteristics, dominate the national or regional production of
selected crops, output price increases may occur when air pollution
reduces crop yields. These price increases, again as explained in
chapter 4, will reduce the well-being of consumers. In addition,
if increases in market prices are insufficient to offset reduction in
output (demand is relatively elastic), producers may also be made worse
off. On the other hand, if demand is relatively inelastic, they will be
made better off. Consumers, however, are always made worse off.
Seasonally (mainly in winter and in spring), Southern California
produces a major share of the nation's vegetables and fruits. Also,
large volumes of field crops such as cotton and sugar beets are grown in
the region. The adverse biological effects on many of these crops of
smog that periodically blankets the region are well-documented.
However, attempts to assess economic impacts of these effects have been
few. Moreover, as explained in chapter 4, those attempts that have been
made simply multiply the estimated reductions in yields by an invariant
price. This method is, as we have seen, especially inappropriate for
crops having geographically concentrated production patterns since their
market prices will vary with the quantity available from the region.
Furthermore, the method is unable to account for changes in cropping
patterns that may be induced in response to pollution. This difficulty
resembles that of standard epidemiology which, as explained in chapter
5, does not account for human economic responses and adaptations to
pollution.
In the research presented in this chapter, a more general and
powerful methodology was employed to assess the economic impact of air
pollution upon fourteen annual vegetable and field crops in four
agricultural subregions of Central and Southern California. These
subregions are shown in the map below.
The study included an analysis of changes in comparative economic
advantages between and among crops and growing locations in response to
increased air pollution. In addition, the method used makes it possible
to distinguish between the impact upon consumers and that upon producers
of these air pollution-induced changes.
61
-------
SOUTHERN
DESERT
agricultural pro-
ducing areas of
study crops (Source
Johnston and Oean,
?g 55) •
monitoring stations
The particular method used in this analysis is called mathematical
programming. This is a type of economic modeling analysis which, given
information about available technologies and about the costs of inputs
and the demands for outputs, can be used to find the maximum value for
an economic objective. For example, in the case of a private
business, this procedure might be used to find that combination of
inputs and outputs that would make the firm's profits a maximum.
For the crops and farming operations in the analysis reported
here, the method was asked to find that combination of crops and outputs
that would not only give maximum profits to the farmer, but that would
maximize the sum of those profits Plus the consumer's surplus obtained
by solutions of the problem--once under the assumption that there is no
air pollution, and then again, under the assumption that the levels of
air pollution prevailing in 1976 existed.
The difference between consumer's surplus plus profit' under the
two circumstances is, then, an estimate of the economic damage, or
inversely, of the benefits of cleaning up from a condition of 1976
pollution to no pollution at all. The needed information about the
links between air pollution and yield and about demand elasticities for
various crops were both obtained from the large published literature
which exists in Southern California on these matters, and through
original statistical analyses by the researchers.
62
-------
For the regions analyzed, the following table presents estimated
air pollution-induced percentage yield reductions for 1976 for the
fourteen crops studied, given the actual 1976 cropping patterns and
locations. Four vegetable crops, broccoli, cantaloupes, carrots, and
cauliflower, displayed no yield effects in these estimates. Reductions
in lettuce yields occurred only in the South Coast, and these effects
were slight. However, lima beans, celery, and cotton suffered
substantial yield reductions, while potatoes, tomatoes, and onions
exhibited moderate losses at observed pollution levels.
Regional percentage yield reductions were by far the greatest in the
South Coast, followed by the Southern San Joaquin, the Southern Desert,
and the Central Coast regions. This ordering of regions by yield
reductions corresponds to how they rate in terms of smoggy conditions.
Region
Crop
Southern
Desert
(1976)
South Coast
(1976)
Central
Coast
(1976)
Southern
San Joaquin
(1976)
Vezetables
Reduction
in 'iield
Beans, processing
green lima
—
15.71
1.57
9.45
3roccoli
—
0.00
0.00
—
Cantaloupes
O.JO
0.00
n.a.
0.00
Carrots
0.00
0.00
o
o
o
0.00
Califlfevar
~
0.00
0.00
—
Celery
—
12.57
1.23
—
Lettuce, head
0.00
0.03
0.00
0.00
Onions, fresh
0.00
1.99
0.40
—
Onions, processing
0.00
1.99
0.40
1.35
Potatoes
—
4.20
0.43
1.95
Tomatoes, fresh
0.00
4.20
0.43
1.95
Tomatoes, processing
0.00
4.20
0.43
1.95
Field
Cotton
9.40
L8.70
n.a.
6.90
Sugar 3eets
0.00
1.63
0.33
1.10
To estimate the extent to which air pollution reduced crop
production in the individual study regions, the 1976 percentage yield
reductions were used to calculate what per acre yields for each crop in
each region would have been if there had been no air pollution.
Given these new per acre yields, the mathematical programming model was
used to calculate new cropping patterns and locations of production, as
well as associated effects on producer profits and consumer's surplus.
The results show that the Southern Desert region would experience
a slight increase in production of most crops susceptible to air
pollution
63
-------
damages, with significant increases in the production of processing
onions and cotton. Those crops more resistant to air pollution damages,
such as carrots and lettuce, exhibit slight declines in. production.
For the other three regions, some crops, such as cauliflower,
lettuce, and broccoli, that are rather tolerant of air pollution, record
minimal changes in production levels. However, broccoli and cantaloupes
in the South Coast region are two exceptions. The very significant
decrease in the production of these air pollution-tolerant crops is due
to their substantially reduced profitability relative to crops that are
more sensitive to air pollution. Production of those air pollution-
sensitive crops, such as lima beans, potatoes, tomatoes, cotton, and
onions, generally tends to increase in each region. As would be
expected, there are only minimal changes in crop production in the
Central Coast region, since 1976 air pollution levels were relatively
small.
We now turn to the central objective of the analysis--estimated
differences in the value of consumer's surplus plus profit "with" and
"without" 1976 levels of air pollution, and the distribution of these
difference among producers and consumers. The following table gives
this information for all the regions combined.
Total Consumer's
Producer
Consumer
Surplus Plus
Producer Profit
$
Profits
$
Surplus
$
With air
1,447,733,227
1,086,788,371
370,944,856
pollution
effects
1,503,024,714
1,122,024,497
381,000,217
Without air
pollution
effects
45,291,487
35,236,126
10,055,361
Estimated losses
due to air
pollution
The results indicate that elimination of 1976 oxidant air
pollution and attendant net increases in aggregate production would have
increased 1976 producer profits by about $35 million and consumer
surpluses by about $10 million, resulting in an increase of about $45
million in the total. This latter figure represents a little under four
percent of the $1.22 billion total farm value of the fourteen crops
produced in the four regions in 1976. About $30.0 million of the
estimated potential increase in the total is due to an improvement in
cotton yields. While this is a significant amount, accepting the
results in chapter 6, it is outweighed by urban damages in the same
region by at least a factor of ten. This result
64
-------
for the most severely polluted major agricultural region in the country
and for the assumption that all pollution is eliminated (probably an
impossibility and certainly ineconomical) suggests the Possibility that
the economic costs of air pollution in the agricultural sector are also
relatively small in the rest of the United States. That this
presumption is not correct is shown in the next chapter where ozone
damage to field crops across the nation as a whole is assessed by the
use of a procedure designed especially for that purpose. The reason is
that the total value of major crops like wheat, corn, cotton, and
peanuts is so enormous that even relatively small reductions in yield
can cause large economic losses for the national as a whole.
65
-------
CHAPTER 10
OZONE DAMAGE TO U.S. AGRICULTURE
The study described in the previous chapter is a rather detailed
look at pollution damage in a single, but very important, agricultural
region in the United States. It was able to incorporate adjustments to
pollution, for example, crop switching, in considerable detail.
In principle, this type of approach could be applied, region by region,
to the entire country. But the resources required to do it would be
considerable and well beyond those available for the project described
here.
Therefore, in the interest of estimating national agricultural
benefits, it was necessary to develop a simpler methodology, that could
use existing data sets. Data limitations laid some restrictions on the
study. The only pollutant considered was ozone, and the only crops
considered were wheat, corn, cotton, soybeans, and peanuts. But ozone
is thought to be the major pollutant affecting agriculture, and these
five field crops account for more than 60 percent of the total value of
U.S. agricultural production.
The methodology developed to assess agricultural damage on a
national scale is called the "Region Model Farm" (RMF) approach.
Essential to this approach is a set of data developed and maintained by
the U.S. Department of Agriculture. The USDA refers to these data as
the "Firm Enterprise Data System" (FEDS). FEDS provides people studying
agriculture with sample operating budgets that describe the entire cost
structure for producing an acre of a particular crop in a specific
region of the continental U.S. The budget is representative of the
average agricultural practice in that region and is verified with a
battery of farm level surveys every two years. A single budget for the
production of soybeans in southeastern North Carolina, for example, may
include cost information on as many as 2 00 inputs to agricultural
production, the average yield per acre to be expected, and the total
number of acres planted in the region. FEDS divides the U.S. into over
200 producing areas. Thus, when the present study examines the cost of
producing wheat, for example, it considers production cost for over 160
regions where wheat is produced in the United States.
The reason why this fine detail on costs is needed is that the
major way in which pollution affects agriculture is through yield
reduction. Since this is so, reduction of pollution will permit a
particular amount of agricultural production to take place at reduced
cost. This cost reduction is one, and the largest, component of
benefits (reduced damage) of pollution control. Thus, if one can
calculate this reduction on a national
66
-------
scale, a major step will have been taken in estimating agricultural
benefits. To do this, total costs of agricultural production must be
calculated before and after pollution control, and since costs of
production vary by region, the fine regional detail provided by FEDS is
needed to do this accurately.
To calculate cost, the study assumed that for each of the FEDS
producing areas, the representative farm budget for a particular crop
type reflects both the cost and yield existing for that budget year, for
given prices of inputs, outputs, and ambient ozone concentrations.
The FEDS budgets are on a per acre basis and can be added up across all
of the planted acres covered by a budget for a particular crop.
Given these data, the aggregate cost of production can be
estimated for whatever the actual output is in a given year. The
procedure used to do this assumes that production is limited by
available land for a particular crop in a given region. All the regions
capable of producing the crop under consideration are then arrayed in
order of increasing cost for the entire country under the further
assumption that each region produces the maximum output that available
land will allow. This latter assumption will be true for all regions
except the highest cost region included, where the maximum output may
not be needed to complete the total output actually produced in a
particular year, say, 1982.
One can illustrate how this works with the aid of a simple graph
that shows the results of this procedure for only the least cost region,
say for corn, and for the next higher one.
Cost of
Producing
Corn $
Unit Cost of Producing
Corn in Next Lowest -»¦
Cost Region
Cost of
Producing
Unit
Total Cost of Producing
Corn in Region Two
Unit Cost x Quantity
Produced
Corn in
Lowest
Cost
Region
Total Cost of Producing
Corn in Region One
Unit Cost x Quantity-
Produced
Q
Total Produced
from Region 1
Total Produced
from Region 2
Quantity of Corn Produced
67
-------
As shown, the total cost of producing corn in the lowest cost
region is the unit cost times the amount produced there, and similarly,
for the next lowest cost region, and so on, until the quantity produced
equals actual recorded national output for that year. Since there are
up to 160 regions that might produce in a particular crop, a graph like
the above, but with all the regions included, would come pretty close
to a smooth curve when seen as a whole. Let us depict such a graph as a
smooth curve and refer to the following diagram.
Cost of
Production
of Corn
/
/ Total
Cost Of
Production
of Corn 1982
Corn Output 1982
To economists, a curve of the above type is known as a "marginal"
cost curve. It displays the increment in total cost for each unit as
output is increased. In parallel with principles discussed in chapter 2
with respect to demand concepts, the area under a marginal cost curve
equals the total cost of producing whatever number of units of output
are produced.
How are these ideas related to estimating the benefits from
reduced ozone? As stated, the major source of benefits (reduced
damages) comes from being able to produce any specified output at lower
cost when pollution effects are controlled. The effect of this is to
shift the marginal cost curve downward as depicted below.
The area B, the difference between the areas under the two curves,
would then be the benefit of reducing pollution to such an extent that
marginal production costs would fall to the lower curve.
We have seen how the upper curve can be calculated with existing
data, but how do we get to the lower curve? To do it, three items of
further information are needed. First, one must know what existing
levels of ozone are in each producing area. Secondly, one must know how
a proposed ozone policy would affect these levels, and third, one must
know how this change in levels would affect yields (the dose-response
relationships we have
68
-------
Marginal Cost of
Production Before
Pollution Control
Cost of
Production
$
Marginal Cost of
Production After
Pollution Control
Q
Quantity of Output
encountered so frequently) . The first two items were supplied by EPA
specialists based on extrapolations from ozone monitoring stations and
on estimated effects of possible pollution control policies.
The third item was estimated by the research team from another
data set available from USDA. These data result from experimental work
conducted by the National Crop Loss Assessment Network (NCLAN).
The approach involves subjecting particular crop varieties to alternate
levels of ozone under laboratory conditions of experimental control.
The relation between yield and ozone concentrations for each crop was
estimated from these data by statistical regression--a technique we have
encountered numerous times in previous chapters.
With these items of information at hand, it is a straightforward
matter to estimate a marginal cost curve corresponding to a new level of
ozone concentration.
But as indicated in chapter 4 and reemphasized in the previous
chapter, if costs are significantly affected by pollution, a price
change will occur, and if the demand for the product is at all elastic,
an associated change in consumer's surplus will take place.
Thus, to get a complete estimate of benefits, one must calculate the
change in consumer's surplus as well as the cost change. In the
following diagram, if the demand curve is as shown, price would fall
from PI to P2. In an industry where there is no significant element of
monopoly; price is determined by the intersection of the marginal cost
line and the demand line. This is because at any price above this,
consumers are willing to pay more than it costs to produce additional
units of output, and at any lower price, additional units of output
cannot be sold for what it costs to produce them.
69
-------
MC
MC
1
2
0
In the diagram, the area 0, x, y is the reduction cost discussed
in connection with the earlier diagram. But the area x, y, z is an
additional benefit which, in accordance with principles discussed
earlier, consists of increases in both consumer's surplus and producer's
profit. To estimate this additional benefit quantitatively, one must
have an estimate of the elasticity of demand for each crop. For
purposes of the study described here, these estimates were taken from
the published literature.
Using these tools, estimates of benefits were made for two
different regimes of ozone change. The first was specified by EPA and
must be regarded as highly unrealistic. The second was developed by the
research team and seems more plausible.
Under the EPA-supplied scenario, it is assumed that for any
hypothetical ozone standard which is to be evaluated, all rural areas
will be exactly at that standard. Since at present particular areas may
be either above or below that standard, under this scenario gains and
losses may cancel out. It is even possible that change from the existing
situation to a tighter standard, given the uniformity assumption, could
results in negative benefits. Since ozone stems mainly from urban
areas, a more realistic assumption would be that if the standard is
tightened in those areas, levels would also fall in the affected rural
areas and not rise anywhere. The latter was assumed in the scenarios
developed by the research staff.
It should also be said that there are very few ozone monitors in
rural areas so that the estimates for those areas are mainly
extrapolations from measurements made in the nearest urban areas.
The accuracy of these extrapolations is very uncertain, thus adding to
the uncertainties inherent in other parts of the estimating procedure.
Interestingly, even under the EPA scenario, a substantial
reduction in ozone concentrations is calculated to yield large benefits.
For example, in the southeastern United States, which is the largest
soybean producing
70
-------
region in the nation, the average rural ambient ozone concentration was
estimated to be about .055 parts per million (PPM) in 1978. The
concentration at urban monitors is usually about twice that estimated
for rural areas, or in this case, about .11 or .12 PPM. The
current national ambient standard is .12. According to the model
calculation, a standard of .05 PPM, if that concentration prevailed
everywhere in rural areas in the region, would yield benefits in soybean
production of more than six hundred million dollars. An extreme
reduction to .01 PPM would yield benefits of more than two billion
dollars in that region alone, according to the model, but such a large
reduction is probably impossible to achieve.
As noted, benefits were also calculated on the basis of more
realistic assumptions about what would happen to ozone levels in rural
areas when the urban ozone standard is tightened--namely, that
concentrations in rural areas would also go down. It was assumed that
alternative ambient standards would apply at monitor sites where ozone
is actually measured. The translation of that standard into ambient
levels in rural areas was then accomplished by means of a very simple
dispersion model (see chapter 3). Otherwise, the methodology was the
same as that described in this chapter and used to estimate the EPA
scenarios. The table below gives some of the results. It was assumed
that the national ambient standard of .12 PPM was met initially at each
monitoring site. This is the base for the calculation of benefits
associated with increasingly strict hypothetical standards. The values
in the table are in 1978 dollars and represent total national benefits
for each crops listed.
03 Con-
centra- „ , „ . .
Peanuts Cotton
tion
PPM
Corn Wheat Soybeans
.09
. 10
.11
. 12
138,482,784
91,143,616
26,731,088
0
224,659,050
154,706,848
73,451,376
0
103,913,808
83,359,392
42,747,408
0
122,513,392
83,394,032
42,794,544
0
1,077,526,832
864,235,888
218,492,344
0
The table shows large national benefits from comparatively small
reductions in ozone concentrations. This is because, even though
effects on yields may not be especially large, the total value of annual
production of these crops is so enormous that even a small yield
response translates into a large number of dollars of benefits.
For example, the total national production (price lines quantity) of
soybeans has in recent years been in the twelve to fourteen billion
dollar range.
As usual, uncertainties pertain to these results. But they do
suggest that agricultural benefits of ozone could be quite large.
71
-------
CHAPTER 11
NATIONAL FRESHWATER RECREATION BENEFITS OF WATER POLLUTION CONTROL
INTRODUCTION
One of the most important pieces of national environmental
legislation created during the 1970S was the comprehensive Amendments of
the Federal Water Pollution Control Act. These amendments, signed into
law in 1972 and further amended in 1977, in reality constituted a major
piece of legislation in their own right, dramatically redirecting the
water pollution control efforts of the nation and setting out ambitious
national goals, expressed both in terms of discharge controls and of
resulting water quality.
Criticism of the amendments and debate over their goals and
requirements began during the legislative process and has continued,
with more or less heat, to the present. One important critical position
is that the goals are too ambitious, that is, the benefits of meeting
the goals (and related requirements) are asserted to be too small to
justify the costs of compliance. This argument over the balance of
benefits and costs can never be resolved entirely by research but the
project described in this chapter was undertaken in the conviction that
it should be possible to improve methods for estimating at least some of
the benefit categories associated with water pollution control. The
particular one addressed is the benefits from recreational fishing in
fresh water bodies of the United States. From the outset the intent was
to design a method for estimating benefits for the nation as a whole
rather than benefits for particular sites. In this respect, it
resembles the study discussed in the previous chapter.
In undertaking this research a primary question concerned the ways
in which water quality improvement would affect fresh water fishing in
favorable directions. Two major ways were identified.
First, it tends to increase the total availability of fishable
fresh water bodies by reducing the incidence of conditions such as low
dissolved oxygen and heavy sediment loads that make it difficult for
fish to survive.
Second, it produces changes in the types of fish that can survive
in particular water bodies. Simply put, clean water means "game" fish
such as trout or bass and dirty water means rough fish such as carp or
buffalo. In general, fishermen prefer game fish. Therefore, pollution
control tends to increase the amount of water yielding high quality
fishing relative to that yielding low quality fishing. Given this
view of the benefit producing mechanism, one can one work toward a
methodology for making national
72
-------
benefit estimates based on it. As explained in Chapters 3 and 4, and
illustrated a number of times since, benefit estimation for
environmental improvement requires the understanding of a number of
linkages. There follows a brief review of them in the context of this
particular study:
(i) how implementation of the law will affect pollution
discharges by location, quantity, and pollutant type
across the entire nation;
(ii) how the prepolicy and postpolicy discharge levels
affect ambient water quality (or how ambient quality
changes as discharges change) in terms not only of such
familiar indicators as dissolved oxygen, but in terms
of supportable fish population types;
(iii) how increases in total amounts of water supporting
recreational fishing and shifts in the composition of
that water toward more highly valued fish species
affects numbers of anglers and the amount of time they
spend fishing.
In addition, one needs to be able to value
(iv) fishing activity of various kinds (i.e., for practical
purposes based on days spent fishing for various
species--rough fish vs. game fish.)
The novelty of this study and its main contribution to
methodological development lies in the ingenious way it was able to link
models together to structure these linkages and how it was able to take
existing and newly developed data sets to estimate them quantitatively.
I turn now to a discussion of each step in the procedure seriatim.
DISCHARGE REDUCTIONS AND LINKAGES TO AMBIENT QUALITY AND FISH
An initial need is an understanding of the "fishability" of the
nation's water prior to the implementation of the Federal Water
Pollution Control Act. A data base was available from the Fish and
Wildlife Service that permitted estimates of fishable water by state
(the state is the basic geographical unit on which this study
operated). But these data do not provide a basis for the breakdown
between rough fish and game fish mentioned above and basic to the
methods used in this study. For this purpose the researchers did
their own survey of state fish and game officials asking them for a
breakdown by species category for their own states. Using these data
they found that for the contiguous 48 states and the District of
Columbia, there are about 30.6 million acres of fishable fresh water
consisting of about 2 0.4 percent cold water game fisheries, 68.4 percent
warm water game fisheries, and 11.2 percent rough fisheries.
To go from this present condition to how fishability would be
affected by implementation of discharge controls requires a knowledge of
the amount and locations of discharges prior to implementation of the
1972 amendments
73
-------
and the same information after implementation. Then one must estimate
how this change will affect ambient conditions in water courses, and
how, in turn, these would affect fishability.
The first three kinds of information were established by the use
of the Resources for the Future's Water Quality Network (WQN) model
described in chapter 3. This model was designed specifically to answer
those questions and it was run for four scenarios representing, albeit
roughly in some cases, stages in the implementation of the law.
In what follows, I will focus on only one of these stages. This is for
simplicity and also because the quantitative benefits results must still
be regarded as rather experimental. The stage of implementation is the
Best Practical Control Technology Currently Available (BPT for short)
requirement that was to be achieved by all point sources of wastewater
discharge by July 1, 1977. This goal may reflect about where we are now
in our current control efforts.
At best, the WQM model provides a reasonable estimate of the
impact of policy changes on one important aspect of ambient conditions:
dissolved oxygen--it does not translate directly into fishability.
Indeed, making that step is an undeveloped discipline.
Accordingly, rather heroic measures were called for. Fortunately, a
fisheries biologist was willing to use his knowledge and skill to make
a survey of the literature to develop a set of rules that appeared to
capture whatever consensus exists on the water quality conditions needed
for the survival and reproduction of fish populations of various types.
These rules were then applied to the results of the WQN model to provide
estimates of the acreage of different kinds of fishing availability by
state, and by aggregation, for the nation as a whole. The results for
BPT for the whole country are shown below:
Baseline and Projected
Fishable water
total
Fishery shares
Warmwater
Coldwater gamefish/
(1000 acres) gamefish panfish Rough fish
Pre-FWPCA/CWA conditions
30,615
0 .204
0 .684
0.112
Pollution control to level
represented by Best
practicable technology
(BPT)
30,721
0 .227
0 .736
0 . 037
74
-------
The reader may be struck by how small the increases in total
fishable water is; only about one hundred thousand acres from a base of
more than thirty million. This is because a very large proportion of
U.S. fresh waters was already fishable before implementation of the
water pollution law. However, at the same time, the water regarded as
unfishable or supporting rough fish only is projected to decline
dramatically. This does not mean a proportionate decline in rough fish
populations, but rather a large increase in the water rough fish will
share with warm and Coldwater game fish.
The next step is to devise ways of converting the water quality
results into changes in fishermen's participation in various kinds of
fishing. I turn to a discussion of that in the next section.
Before proceeding, however, it is pertinent to note that what has
been discussed so far are not types of research and modeling that are in
the usual purview of economics. But the situation is reflective of the
fact, as we have also seen in other chapters, that existing models of
natural systems rarely fit the needs of the economist who would estimate
the benefits of environmental improvement. Accordingly he is often
forced into disciplinary imperialism.
BEHAVIORAL ECONOMIC ASPECTS OF THE STUDY
I now turn to steps in the analysis that are more clearly economic
in character. To finally wind up with an estimate of total activity in
various types of fishing, the individual fisherman's chain of decision
about recreational fishing is broken down into several logical stages.
The first choice is whether to do any fishing at all. The
hypothesis adopted in the research on this question is that the decision
of whether to fish or not is sensitive, among other things, to the
opportunity to fish, represented by the quantity of fishable water. The
object of this first stage of the research is then to quantitatively
estimate how the decision to fish is influenced, in the population at
large, by the availability of fishable water. Regression analysis is
the method used to try to sort out from the data the separate influences
of availability of fishing opportunity and those other influences that
might affect the decision (e.g. income, sex). Regression analysis is
the basic econometric tool we have encountered so many times in these
pages and which is briefly explained in chapter 5.
The indicators of existing fishable water are the state level
estimates, already mentioned above, divided by the state population to
get a per capita measure. This is rather crude but a more refined
indicator was not available.
The other needed data for this stage of the research were obtained
from a very large survey (more than 300,000 individuals) conducted by
the Fish and Wildlife Service in 1975. This was a telephone survey and
its primary intent was to determine whether or not individuals
participated in
75
-------
hunting, fishing, and other recreational activities associated with
wildlife. The survey also contained information on other pertinent
variables such as age, sex, income, etc. so that it was Possible to
include them in the regression analysis and control for their Possible
effects on participation. The dependent variable was the decision to
fish or not to fish. Since the measure of fishable water availability
was included among the independent variables, once the coefficients of
the equation are estimated, the size of the availability variable can be
changed and the corresponding change in participation calculated. We
have seen regression analysis results used in a similar way in other
chapters of this volume. For example, in projecting the effect of air
quality improvement in chapter 8.
So far, all the analysis permits us to do is to project fishing in
general as a function of water quality. But since, as I have indicated,
different types of fishing (warm water game fishing, cold water game
fishing, and rough fishing) probably differ in value we must also be
able to project how likely a representative individual is to pick each
of these types if he does decide to fish.
For this purpose data developed by the Fish and Wildlife Service
for a subsample of the large telephone survey mentioned earlier could be
used. A mail questionnaire was sent to more than 50,000 persons who had
declared themselves to be hunters and fishermen in the large sample.
For this subgroup detailed information was gathered on their
participation patterns, socioeconomic characteristics, and preferences.
Data for the fishermen only was used in analyzing the second stage in
the decision chain, namely once a person has decided to fish how likely
is he to participate in each of the three types of fishing given the
availability of water suitable for each type. While the analysis of
the decision also uses regression techniques, a very complicated
mathematical model involving simultaneous equations had to be used, and
it is not Possible to explain it, even in general, in a book intended
for a nontechnical audience. Suffice it to say that a means was
developed for predicting the likelihood of different types Of fishing
given the availability of different types of fishing water and given
that one had decided to fish at all.
The final stage in the decision chain is the decision that once a
fisherman has decided to engage in a certain type of fishing, how much
time (many days) he will spend in that activity? The same set of mail
survey data was used in the analysis of this question, but once again I
must ask the reader to make a leap of faith and not inquire how.
But the drift of the analysis is now clear. The steps are as
follows: the amount of increase in total fishable water and fishable-
type water associated with water pollution control is given for the
nation as a whole from the models of the previous sections. Given this,
the results of stage 1 are used to calculate how much fishing
participation will increase in general. Then, the results of stages 2
and 3 are used to calculate how this increase in participation will be
distributed across the fishing types and how many days Of increased
fishing of each type will occur nationally
76
-------
as a result of the pollution control policy. This process, and some re-
sults, is laid out in the table below:
Water Quality
Base BPT
Probability of being a fisherman 0.2793 0.2794
Total fishermen (10 6 ) 59-16 59.18
Probability of doing some:
Coldwater gamefish fishing 0.3708 0.3931
Warmwater gamefish fishing 0.6840 0.6776
Rough fishing 0.3499 0.3536
Days per angler per year of:
Coldwater gamefish fishing 13-76 13-73
Warmwater gamefish fishing 18.22 18.49
Rough fishing 10.14 10-55
Total days per year of:
Coldwater gamefish fishing (106 ) 301.8 319.3
Warmwater gamefish fishing (10 6 ) 737.4 741.4
Rough fishing (10) 209.8 220.8
The final problem confronted by this research on the benefits from
improved fresh water fishing opportunities is how to assign dollar value
benefits (willingness to pay) to the increase in each category of
fishing activity. The approach adopted was to estimate a demand curve
for fishing days for each category and to use those to calculate
consumers surplus. The travel cost method, described in general terms
in chapter 2, was the technique selected. I now turn to a brief
discussion of how it was applied in this study.
Recall that the basic assumptions of the travel cost method is
that higher costs of access as reflected in distance from a recreational
site will have the same effect on visitation as an equivalent emissions
fee assuming zero distance from the site. In chapter 4, I
presented a very simple example of how this relationship is used to
develop a demand curve by assuming successively higher admissions fees
and using information on access costs to estimate their effects on
visitation. This establishes points on a demand curve, i.e., the
relationship of price to the number of visitor days. The area under the
demand curve, by principles discussed in
77
-------
chapter 2, is the total willingness to pay of participants for the total
number of visitor days to the site, say a trout fishery. If
one then divides the number of visitor days into this number, one
obtains the average willingness to pay in fishing for trout. The
researchers who conducted the study collected data from a large number
of fishing sites around the country which permitted them, by statistical
means, to make exactly such a calculation yielding average willingness
to pay per Visitor day for each type of fishery.
We are now at a point where a national benefits estimate can be
made. The point of all the earlier machinations was to derive an
estimate of how many days of increased recreational fishing of each type
would correspond to the water quality changes resulting from a reduction
of waste water discharges corresponding to the implementation of a
pollution control policy. Having these numbers in hand, it is a
simple matter to multiply them by average willingness to pay for a day
by fish type and get a total benefit number for freshwater fishing in
the United States. When this is done the following results are obtained
for BPT:
Total Benefits Over Base
Valuation Base (Millions of 1980 Dollars)
Low 3 07
High 683
A few words of explanation are needed about the difference between
the low and the high estimate. For the low estimate travel cost is
figured based on only out of pocket expenses--gasoline, restaurant
food, motels, etc. This is the conventional method. The higher
estimate takes account of the fact that the fisherman may also attach a
cost to the time it takes to get to the site. For the higher figure an
estimate of this Cost is made by attaching average wage rates to the
travel time needed to get to the site.
Needless to say large uncertainties attend these numbers and, as
already said, they must be regarded as largely experimental.
Nevertheless, in view of the heavy costs of the national water quality
improvement program they may strike the reader as being quite low.
There are several things to be said in this connection. First, the
reader should recall that in terms of the availability of fish species
the vast majority of the nation's fresh water was already fishable prior
to the 1972 amendments. Secondly these estimates are partial in the
sense that they consider only the fresh waters of the United States, and
even then they do not include as values that may be accrued to fisherman
the possible effects of pollution control on the aesthetic aspects of
the fishing experience. At present, the search is underway to extend
the methodology developed in this study to effects of pollution control
on marine (salt water) recreational fisheries.
78
-------
CHAPTER 12
A SURVEY RESEARCH METHOD FOR ESTIMATING NATIONAL
WATER QUALITY BENEFITS
INTRODUCTION
The research reported in the last chapter was designed to yield
national recreational fishing benefits of water quality improvement.
But its basic approach was still to use subregions as units of analysis
and to aggregate by adding up the results. In this sense it was still
"site specific," although less so than, say, the visibility study
reported in the next chapter. Thus it can be described as a large scale
simulation falling somewhere between a particular site (or micro) study
and a national survey that asks respondents directly about their
willingness to pay for national programs of pollution control. This
last procedure we have called the "macro" approach. Among other
potential advantages of such an approach, two are especially important.
First, a randomized national sample of persons can be interviewed which
permits well-established statistical procedures to be used to
extrapolate the results to the entire population. Second, one can
inquire about "intrinsic" or existence benefits as well as user
benefits.
The second reason invites a bit of explanation. Because the U.S.
population politically supports very expensive programs of water
pollution control, much more costly than the benefits estimated for
recreational users in the last chapter for example, the researchers were
led to believe that there must be some form or forms of benefits
accruing to persons who do not actually use particular water bodies.
We termed such benefits variously as intrinsic or existence benefits.
These benefits may accrue because persons value the options for possible
use that are opened to them when water bodies are cleaned up. This
type of value has been discussed widely in the economics literature and
has come to be called option value. Other intrinsic values may accrue
from a sense of national pride or rectitude associated with having clean
waters. One of the main conclusions of the research reported in this
chapter and in the following one, which as mentioned deals with air
quality, is that intrinsic benefits definitely exist with respect to
environmental improvements or maintenance. Moreover, and with the usual
caution about accuracy of results, not only do they exist, but they are
large, perhaps larger than user benefits in some instances.
Some aspects of the water quality situation made it more appealing
for an experimental application of the macro approach than is air
quality. Chiefly, goals of our national policy are set out in a manner
that would
79
-------
let most of the population understand what they mean in terms of
ordinary experience. The objectives are stated to be to make all the
nation's water fishable and swimmable in successive stages.
Furthermore, most of the cost of these programs is to be paid from taxes
levied at the national levels so that respondents can be realistically
asked how much in added tax burden they are willing to pay for improved
water quality across the whole nation. Neither one of these situations
holds with respect to air quality, so it would be much harder to pose
understandable and realistic alternatives in a national clean air
survey.
A macro study, then, is potentially useful for doing a benefit -
cost analysis for whole national water programs. It should be noted,
however, that it is not a substitute for site-specific studies in other
applications. For example, determining whether or not the benefits
outweigh the costs of a water quality improvement program in the Potomac
Estuary would require a site-specific study.
RESEARCH PROCEDURES
One problem with national surveys is that they are quite
expensive. What made it possible to conduct an experiment with the
macro approach, given available resources, was that the researchers were
able to piggyback some water quality questions onto a survey being
funded by another source. After the interview for the other survey was
completed, the interviewers administered a sequence of benefits
questions that had been carefully pretested by researchers on the
benefits project. From the respondents' perspective, the two
interviews appeared as one long interview. In all, 1,576 personal
interviews of a national probability sample of persons eighteen years of
age and older were completed. The sample was designed and the
interviews were conducted by Roper and Cantil.
A penalty of this add on approach proved to be that an
unfortunately large number of persons failed to complete all of the
questions. In part this was because they came at the end of an already
fairly lengthy survey and in part because it was not possible to
undertake special training of the interviewers to administer the
benefits section. Because of the likelihood of item response bias
(caused by respondents failing to answer individual items), the
researchers regarded their estimates as only suggestive and warn against
regarding them as definitive. The main intent of the experiment was not
to develop definitive estimates at this stage but to test whether a
macro approach is applicable to water quality benefits investigation.
The low response rate presumably could be cured by an improved
questionnaire and by training of the interviewers. A study is currently
being planned in which both of these elements will exist.
80
-------
WATER POLLUTION LADDER AND VALUE LEVELS
The levels of water quality for which the research team sought
willingness to pay estimates are "boatable," "fishable," and
"swimmable." These levels were described in words and depicted
graphically by means of a "water quality ladder." Use of these
categories, two of which are embodied in the law mandating the national
water pollution control program, permitted avoidance of the
communications problems associated with description of water quality in
terms of the numerous abstract technical measures of pollution (oxygen
depletion, for example). Although the boatable-fishable-swimmable
categories are widely understood by the public, they did require further
specification to ensure that different people perceived them in a
similar fashion.
Boatable water was defined in the text of the question as an
intermediate level between water which "has oil, raw sewage and other
things in it, has no plant or animal life and smells bad" on the one
hand, and water which is of fishable quality on the other. As discussed
in the previous chapter, fishable water covers a fairly large range of
water quality. Game fish like bass and trout cannot tolerate water that
certain types of fish such as carp and catfish flourish in. In
pretests, experiments were made with two levels of fishable water--one
for "rough" fish like carp and catfish, and the other for game fish like
bass--but a single definition of "fishable" was adopted as water "clean
enough so that game fish like bass can live in it" under the assumption
that the words "game fish" and "bass" had wide recognition and connoted
water of the quality level Congress had in mind. Swimmable water
appeared to present less difficulty for popular understanding since the
enforcement of water quality for swimming by health authorities has led
to widespread awareness that swimming in polluted water can cause
illness.
Because willingness to pay questions have to describe in some
detail the conditions of the "market" for the good, they are inevitably
longer than the usual survey research questions. Respondents quickly
become bored and restless if material is read to them without giving
them frequent opportunities to express judgments or to look at visual
aids. The questionnaire for this experiment was designed to be as
interactive as possible by interspersing the text with questions which
required the respondents to use the newly described water quality
categories. They were also handed the water quality ladder card which
was referred to constantly during the sequence of benefits questions.
The following figure shows the card. The top, step 10, was called
the "best possible water quality," and the bottom, step 0, was the
"worst possible water quality." The card is "anchored" by designating
five levels of water quality at different steps on the ladder. Level
E, at .8, was specified as a point on the ladder where the water was
even unfit for boating. Level D, 2.5, was where it became okay for
boating; C at 5 was fishable, B at 7 was swimmable, and 9.5 was
identified as A, where the water is safe to drink.
81
-------
OUTER QUALITY LAODEX CAM)
BEST POSSIBLE
WATER QUALITY
10
i a SAFE TO DRINK
A
SAFE FOR SWIMMING
(b
H.
_L| c GAME ?TSH l:k.
!
-------
WILLINGNESS TO PAY QUESTIONS AND ANSWERS
Questions about willingness to pay should seem realistic to
respondents. Accordingly, they were couched in terms of annual
household payments in higher prices and taxes because this is the way
people do pay for water pollution control programs. A portion of each
household's annual federal tax payment goes toward the expense of
regulating water pollution and providing construction grants for sewage
treatment plants. Local sewage taxes pay for the maintenance of these
plants. Those private users who incur pollution control expenses, such
as manufacturing plants, ultimately pass much or all of the cost along
to consumers in higher prices. Thus, this payment method has a true
ring for the respondents.
As explained in chapter 4, "starting point bias" can be an
important problem in bidding games and surveys. That is, a high
starting bid from an interviewer may elicit a higher bid from a
respondent than a low starting bid. A major methodological innovation of
the research reported in this chapter is the development of a device for
eliminating such a bias, the "payment card."
In this technique, the respondent is given a card which contains a
menu of alternative amounts of payment which begin at $0 and increase by
a fixed interval until an arbitrarily determined large amount is
reached. When the time comes to elicit the willingness to pay amount,
the respondent is asked to pick a number off the card (or any number in
between) which "is the most you would be willing to pay in taxes and
higher price each year" (italics in the questionnaire) for a given level
of water quality. Thus, the interviewer suggests no bid at all.
It turns out, however, that this presents some problems of its
own. In initial pretests, it was found that the respondents had
considerable difficulty in determining their willingness to pay when a
card was used which only presented various dollar amounts. A number of
them expressed embarrassment, confusion, or resentment at the task, and
some who gave amounts indicated they were very uncertain about them.
The problem lay with the lack of benchmarks for their estimates. People
are not normally aware of the total amounts they pay for public goods
even when that amount comes out of their taxes, nor do they know how
much they cost. Without a way of psychologically anchoring their
estimate in some manner, they were not able to arrive at meaningful
estimates. They needed benchmarks of some kind which would convey
sufficient information without biasing their responses. The most
appropriate benchmarks for willing to pay for water pollution control
would appear to be the amounts they are already paying in higher prices
and taxes for other nonenvironmental public goods. Amounts
were identified on the card for several such goods and further pretests
were conducted. These showed the benchmarks made the task meaningful
for most people.
But the use of payment cards with benchmarks raises the
possibility of introducing its own kind of bias. Are the respondents
who gave amounts for water pollution control using the benchmarks for
general orientation or are they basing their amounts directly on the
benchmarks themselves in some
83
-------
manner? In the former case, people would be giving unique values for
water quality; in the latter case, they would be giving values for water
quality relative to what they think they are paying for a particular set
of other public goods. If the latter case holds and their water
quality values are sensitive to changes in the benchmark amounts or to
changes in the set of public goods identified on the payment card, their
validity as estimates of consumer surplus for water quality are suspect.
Tests for this kind of bias were conducted in the pretest by using
different versions of the payment card. No bias was found, and so the
"anchored" payment card was deemed to be a suitable device for the full-
scale experiment.
Tests were also conducted to attempt to discover if any of the
other sorts of bias discussed in chapter 4 were inherent in the
questionnaire. Again, none were found.
A final point on the payment card. What people actually pay f or
publicly provided goods varies with their income. To correct for this,
four different payment cards were developed corresponding to four income
classes. At the appropriate point in the interview, the interviewer
gave the respondent the payment card for his or her income category
which had been established by a prior question.
As already discussed, the respondents valued three levels of water
quality which were described in words and depicted on the water quality
ladder. They were first asked how much they were willing to pay to
maintain national water quality in the boatable level.
Subsequent questions asked them their willingness to pay for overall
water quality to fishable quality and swimmable quality. The average
willingness to pay amounts given by the respondent for the two higher
levels consists of the amounts they offered for the lower levels plus
any additional amount they offered for the higher level.
The average annual amounts per household (1981 dollars) for those
respondents who answered the willingness to pay questions turned out to
be:
Boatable $152
Fishable 194
Swimmable 22 5
The most substantial benefit is for boatable water. The
respondents are willing to give about 2 0% more for fishable water than
boatable water, but only about 15% in addition to make the water
swimmable. As we will see later, these are large amounts.
The data also permitted making a rough distinction between the
recreation and the intrinsic values discussed earlier. Since the
willingness to pay questions measure the overall value respondents have
for water quality, the amount given by each respondent represents the
combination of recreational and intrinsic values held by that person.
But it was possible to tell from the questions whether or not a person
actually engaged in water-based recreation. It was reasoned that the
values
84
-------
expressed by the respondents who do not engage in in-stream recreation
should be almost purely intrinsic in nature. In calculating the
average willingness to pay amount for the nonrecreators alone,
therefore, we get an approximation of the intrinsic value of water
quality. By subtracting this amount from the total the recreators are
willing to pay, one can estimate, in a rough way, the portion of the
recreators' benefits which are attributable to intrinsic values.
When this is done, it is found that intrinsic value constitutes
about 45% of the total value for recreators, 100% for the nonrecreators
(of course), and about 55% for the sample as a whole. If this is
a correct reflection of reality, it is a major finding and may have
large implications for the future study of benefits from environmental
improvement. This matter will be pursued further in the next chapter,
which deals with visibility in the national parks.
It was noted earlier that, while the sample of persons interviewed
was initially chosen to be random, quite a few respondents failed to
give useable answers. Any aggregate national benefit estimate based on
these data could not therefore be put forward as accurate. Therefore, I
make such an estimate simply to illustrate that the results of this
experiment imply very large values.
There are about 80 million households in the United States.
Assume that the sample results imply an annual willingness to pay of
$200 per household to have high quality recreation waters throughout the
country. This would imply a total willingness to pay of $16 billion.
According to results explained earlier, this would divide about equally
between user and nonuser values. At first this might seem quite out of
line with the value of well under a billion dollars calculated for
recreational fishing in the last chapter. But this is not
necessarily the case. Recall that that estimate is for a relatively
small increase in the nation's fishable waters and that the estimate
from the national survey is the value people attach to making and
maintaining the whole of the nation's fresh waters of high recreational
quality.
But the objective of this experiment was not to produce an
accurate estimate of national benefits, rather it was to test the
feasibility of using a macro approach to the estimation of water
quality benefits. In that, it succeeded.
85
-------
CHAPTER 13
THE VALUE OF VISIBILITY IN THE NATIONAL PARKS
The first case reported in this volume that involves primarily a
preservation issue instead of an amelioration one is visibility in the
national parks. Historically, Americans have placed a high value on
good visibility, that is, the ability to see distant objects clearly.
This yearning for the appreciation of atmospheric visual clarity is
evidenced in the country's early literature and art, including the
journals of Lewis and Clark as well as the masterpieces of the great
American landscape artists of the 19th century. Today that love of
visibility is demonstrated not only by the millions who flock each year
to our Western parks, but also in the high prices brought by those
artists' works of a century ago and by the interest in Ansel Adams'
simple, yet dramatically clear, black and white photographs of Yosemite
and other wonders of the U.S. National Park Service.
Over the past 100 years, Congress has acted to preserve many of
our nation's natural wonders. It did so by creating and by continually
expanding the National Parks, National Wilderness Areas, National
Monuments, National Recreation Areas, and Wild and Scenic Rivers.
Since the 1950s, there seems to have been an increasing concern
that this beauty is threatened by industrial development and population
growth. Pollution from coal-fired power plants became a special
concern with the advent in 1963 of the first unit of the Four Corners
Power Plant near Farmington, New Mexico. It produced a plume that could
be seen clearly for many miles, reducing the clarity of the Visual
experience in areas of northwestern New Mexico, southeastern Utah,
southwestern Colorado, and northeastern Arizona.
By the later 1960s and the early 70s, smog began to appear in
Yosemite Valley on warm summer days. Battles erupted over proposed
coal-fired power plants on the Kaiparowits Plateau and near Capitol Reef
National Park, both in southern Utah, because of their possible effects
on visibility. The increased publicity and concern resulted in
magazine and newspaper articles decrying the loss of visual clarity,
particularly in the western United States and precipitated political
pressures in Congress for legislative steps to protect visibility.
Those pressures culminated in the August 1977 adoption by Congress of
the nation's first specific visibility protection requirements for
national parks and national wilderness areas. One of the large
issues raised by these developments is whether the value of visibility
protection outweighs the cost, including both air pollution control
equipment and the regulatory system. The study reported in this
86
-------
chapter was designed to improve our ability to measure the benefits of
visibility and to provide some actual preliminary estimates of the value
of that visibility in several major national parks and for the region in
which they are located. The region and parks located in it are shown in
the map below. We refer to this as the Grand Canyon Region.
<7/
ICO
*tG20NAl iNOUSTftfAl JAOUFU
U* * m 'wnm
Visibility is the ability to see both color and detail over long
distances. Human perception of visual air quality is associated with
the apparent color contrast of distant visual targets. As contrast is
reduced, a scene "washes out" both in terms of color and in the ability
to see distant detail.
What, then, is the nature of the preservation value of visibility?
That value has at least two Possible components.
First, a scenic resource such as the Grand Canyon attracts large
numbers of recreators. The quality of the experience of these
recreators depends in great part on air quality, in that scenic vistas
are an integral part of the Grand Canyon "experience." Accordingly,
air quality at the Grand Canyon is valuable to recreators. We might
call this economic value, or willingness to pay by users for air quality
at the Grand Canyon, user value. Thus, recreators in the National
Parklands of the Southwest should be willing to pay some amount to
preserve air quality for each day of their own use if their recreation
experience is improved or maintained by good air quality.
The second component of preservation value we have termed
existence value. This concept was introduced in the abstract, in chapter
4, and
87
-------
explained in a more specific context in the last chapter. Individuals
and households which may never visit the Grand Canyon may still value
Visibility there simply because they wish to preserve a national
treasure. Visitors also may wish to know that the Grand Canyon retains
relatively pristine air quality even on days when they are not visiting
the park. Concern over preserving the Grand Canyon may be just as
intense in New York or Chicago as it is in nearby states and
communities. Thus, preservation value has two additive components, user
value and existence value.
During the summer of 1980, over six hundred people in Denver, Los
Angeles, Albuquerque, and Chicago were shown five sets of photographs
depicting both clear conditions and regional haze, each set consisting
of five photographs of a national park vista with different visual air
quality of a general nature; that is, generally increased haziness. The
vistas are from Grand Canyon, Mesa Verde, and Zion. Summer was chosen
for the survey because it is the season of peak visitation.
These photographs were placed on display boards as full frame 8 x
10 inch textured prints, arranged from left to right in ascending order
of visual air quality, with each vista a separate row. An example for
the Grand Canyon is shown below.
WORST
BEST
OESERT VIEW 9 AM
WORST
8EST
TRUMBULL MT. 9 AM
WORST
BEST
TRUMBULL MT. 3 PM
The participants were asked how much they would be willing to pay
for visibility as shown in the five sets of photographs, from worst to
best. They were also asked about their willingness to pay to prevent a
plume from being seen in a pristine area. Two photographs were used in
this connection, one with and the other without a plume. The
photographs were taken from Grand Canyon National Park at the Hopi fire
tower observation point and toward Mt. Trumbull. They were both taken
at 9 a.m. , so the
88
-------
lighting on the Canyon wall and other features is the same. Both
photographs have the same light, high cirrus cloud layer in the sky.
The plume is a narrow gray band crossing the entire vista in the sky,
except where it is in front of the top of Mt. Trumbull. The source was
not industrial or municipal pollution, but a controlled burn in the area
around the Grand Canyon. However, the effect was comparable to what a
large industrial source might produce.
The bidding game based on these photographs reveals the
household's willingness to pay for preserving or improving the degree of
visibility in specific locations of the National Park area described
earlier. The bids offered by interviewees in the preservation value
section of the survey encompasses both pure existence value and user's
valuation of preserving visibility. Since the results did not
permit a completely clean distinction between the two types of bids,
further discussion will concentrate on the preservation value section of
the survey.
The benefits derived from the interview results can be
extrapolated to populations larger than that in the sample (the sample
was chosen in as random a manner as practical) by applying statistical
techniques to the results of the survey. The amount of bids offered by
interviewees to preserve or improve visibility is related to such
factors as income, education, and other personal characteristics.
These relationships can be quantified using the regression type of
analysis that has been explained earlier. After this is done, it is
possible to estimate the value of benefits to residents of the whole
Southwest region as well as the entire nation. This is done by
substituting the average values for these characteristics for each state
into the relationship established by the regression analysis technique
described previously and calculating what the average value of the bid
of a person in that state would be. This value can then be multiplied
by the population of the state as a whole to get a total bid.
When the analysis is performed for the southwestern United States
(for residents of California, Colorado, Arizona, Utah, Nevada, and New
Mexico), the following values are obtained. The figures are willingness
to pay for preserving present average conditions (middle picture) to the
next worse condition as depicted by the pictures and to prevent plume
blight. As the table indicates, the aggregate benefits for the
Southwestern region from preserving visibility in the Grand Canyon
National Park, the Grand Canyon region, and for avoiding a visible plume
over the Grand Canyon is:
Benefits for Total ($ million)
Grand Canyon 4 7 0.0
The Grand Canyon Region 889.0
The Plume 37 3.0
89
-------
about $470, $889, and $373 million, respectively. To estimate the
aggregate national benefits from preserving visibility, a similar
analysis is done for the entire U.S., but additional survey data from
Chicago are included, and the following values are obtained.
Benefits for Total ($ million)
Grand Canyon 3,370.0
The Grand Canyon Region 5,760.0
The Plume 2,040.0
These figures even though their accuracy is highly uncertain, imply that
very large existence values characterize the areas in question.
However, some very recent and highly preliminary experiments with
surveys imply that these figures may be much too high. This matter is
taken up again in the concluding chapter.
Several other observations on the outcomes of the analysis of the
actual interview results are worth mentioning. First, in the
conventional view of the demand for environmental quality, there is a
smooth tradeoff between higher successive levels of environmental
quality and economic benefits, with successive units commanding less
incremental willingness to pay. This is embodied in the depiction of
demand curves in chapter 2, viz.
$
Demand Curve
Improved Environmental Pristine
Quality Conditions
The survey respondents, however, placed a much higher value on a
small initial diminution in visual clarity than on comparable subsequent
decreases. This would produce a very unusual demand curve, resembling
what mathematicians call a step function, something like the following
figure.
Second, again somewhat contrary to expectations, neither past nor
prospective visits to the Grand Canyon Region were shown to be important
90
-------
Pristine
Conditions
Improved Environmental
Quality
Improved Environmental
Quality
determinants of preservation value. On the average, those who had never
seen the Canyon valued it as highly as those who had.
Third, once more unexpected, distance from the region had no
significant relationship to the size of household bids. When corrected
for income and other differences, people in Chicago bid fully as high as
those closer by. However, preliminary further investigation suggests
that this result may not be very robust, being sensitive, for example,
to the sequence in which people are asked about their valuation of
various public goods. Further investigation is clearly indicated, and
the matter is discussed further in the concluding chapter.
Because the Grand Canyon is the dominant feature in a region with
many visitor attractions, one must be especially cautious in extending
these preliminary findings to other recreational attractions. It
seems likely that there are only a very few natural phenomena in the
United States about which Americans have such strong feelings.
Obvious candidates for this short list, would be Old Faithful (in
Yellowstone National Park), Niagara Falls, and perhaps a few others.
The main conclusion of this study is that the magnitude of the
annual benefits when aggregated across households is impressive. While
these are necessarily rather crude extrapolations, the survey results
suggest that Americans place great value on the preservation of air
quality in the Grand Canyon region, and that this valuation is not
necessarily localized in the Southwest. Further, the survey results
suggest that pure existence value may overwhelm a substantial user
value for the national parks in the region.
91
-------
CHAPTER 14
BENEFITS FROM CONTROLLING ACID RAIN
As indicated in chapter 3, acid rain and other mechanisms for the
dispersion and deposition of acid formed from sulfur and nitrogen
emitted from various sources are complex and ill-understood phenomena.
In addition, methods for estimating the economic losses resulting from
damages or economic benefits of prevention of acid rain are not well
developed, nor was it possible within the scope of the project described
here to make much progress in developing them.
Consequently, since the estimates of benefits made for controlling
acid rain are very crude and of no particular interest in terms of
methods development, our discussion here will be very brief. It
is included primarily because of current intense interest in the
phenomenon, and the analysis that was done provides some guidance
concerning directions for future research. The acid deposition problem
among all the areas covered in this volume is perhaps the one most
crying out for additional methods development and improved estimates.
Let us turn first to possible effects on the activities of
agriculture and forestry.
Increases in soil acidity can have a negative effect on the yields of
certain field crops. But it appears that this could be offset by
modest increases in liming operations which already occur for acid-
sensitive crops and that the benefits of controlling acid rain for
this purpose would therefore be small. It is known that there can
also be direct damage to the plant from acid deposition on leaves,
flowers, and fruits, but there is virtually no basis for estimating
the amount of such an effect.
As far as forest growth is concerned, as in the case of
agriculture, there can be both indirect, through the soil, and direct
effects. Indeed, again as in the case of certain field crops, there may
even be short run favorable effects as the acid dissolves plant
nutrients and makes them more available to the trees. But the longer
term effect of this would be reduced soil fertility and slower tree
growth. If some strong assumptions are made, an estimate can be made of
damages resulting from retarded growth. If one assumes, and there is
some evidence pointing in this direction from Swedish studies, that acid
rain would reduce timber growth in Minnesota and east of the Mississippi
(the area of the country thought to be most affected by acid) by five
percent annually, the reductions in yield would decrease the worth of
timber production about six hundred million dollars per year. Assuming
other services of forests, such as
92
-------
watershed protection, fishing, and hunting, were to also be reduced by
five percent, and based on crude estimates by others of the possible
overall value of these services, the total damage including timber and
other services might come to about one and three quarters billion
dollars. This is a substantial sum, but not very large relative to the
costs of controlling acid deposition.
There might also be effects on human health, say by the acid
dissolving and mobilizing heavy metals so that larger concentrations
would get into drinking water or the human food chain. The present
state of knowledge does not permit even very crude estimates to be made
of this possibility. Higher acidity in municipal and industrial water
systems might also result in increased erosion in piping, appliances,
cooling systems, etc. But the adjustment of acidity in such systems, by
the use of lime, is a routine operations and can be accomplished at
small cost.
The big danger in water courses appears to be to those features of
the aquatic ecosystem itself which mankind values. Acid conditions in a
water course tend to destroy the small plants and animals (plankton),
that are the initial links in the fish food chain, and this has a
negative effect on fish population. But the primary way in which fish
populations are destroyed is different. As noted earlier, acid in water
bodies tends to mobilize heavy metals and increase their concentration
in the water. The reproductive capacity of many species of animals,
including fish, is adversely affected by the presence of excessive
amounts of heavy metals. Thus, for a time, as fish numbers decline, the
ones that remain increase in size as competition for food declines, but
then rather abruptly there are
none left. This, of course, destroys commercial and recreational
fisheries. The value of fish taken by commercial fresh water fishing
in the United States is not very large, so the loss there, at least as
measured by present market prices, would not be very great.
The value of fresh water recreational fisheries is, on the other
hand, relatively enormous. Let us make the extreme assumption that all
the recreational fisheries in Minnesota and other areas east of the
Mississippi would totally disappear. If we then take estimates of
willingness to pay for fishing from other studies, it appears that the
loss could, at an Outside limit, be on the order of ten billion dollars
per year in 1979 prices. Additional losses would be caused by the
decline of terrestrial and aquatic animals (other than fish) who are
partly or wholly dependent on the aquatic food chain--certain species of
water fowl, for example.
The other area where our study suggests really major damages might
occur is deleterious effects on materials. As indicated in chapter 3,
acid corrodes metals, eats away at limestone, is harmful to paints and
other coatings and finishes, and damages cloth. Given the huge number
of such items which exist and are exposed to the atmosphere, it is not
very surprising that benefits from protecting them might be large.
Again, in Sweden, where the problems of acid rain first received
widespread attention (because of prevailing winds, Sweden gets inputs of
sulfur and nitrogen compounds from the Ruhr, the Rotterdam petrochemical
complex, and Great Britain), a study has been made of per capita damages
of corrosion and
93
-------
soiling. If one makes the assumption, once again very gross, that this
same estimate can be applied to all persons dwelling in Minnesota and
east of the Mississippi, one gets an annual benefit of avoiding acid
rain of about fourteen billion dollars.
Putting together the various dollar estimates (agriculture, 1-3/4
billion; aquatic ecosystems, 10; and materials damage, 14), one gets
benefits of preventing acid rain in 1978 dollars of about twenty-six
billion dollars--a hefty number indeed. But as stated, many extreme
assumptions were made in generating these numbers, and they are no doubt
too high by quite a lot. An educated guess by the research team was
that the actual figure is probably not more than five billion dollars
for a condition that is characterized by severe effects in the entire
eastern United States.
Of course, even this number cannot be taken very seriously,
because even if it were correct, in all other respects it neglects the
large adjustments in demand and supply which would accompany the types
of changes contemplated. An approach much more like that described for
agriculture study in the South Coast Basin in chapter 8 would be
appropriate in a more sophisticated study.
Perhaps the greatest utility from the acid rain benefits study to
this point is to give some perspective on where the greatest potential
benefits of protection from acid rain are likely to lie. These are, in
rough order, materials damage, aquatic ecosystems effects, and
agriculture and forestry. These categories of damages certainly merit
further study.
But progress in economic research on these questions is highly
dependent on improved dose-response relationships. As indicated in
chapter 4, the relation between intrusions of acid into a water course
and the path to ultimate effects appears to be extraordinarily complex,
including possible sudden changes after a period of time when everything
appears all right on the surface, and the difficulty, if not
impossibility, of reversing them after they have occurred. Ecologists
place great value on diversity of species as an indicator of a
healthy, stable ecological system. Acidification of streams is known to
reduce diversity. But it is not well understood how this ultimately
affects characteristics of the stream that man values. This problem
seems ripe for joint work between economists and ecologists.
All three of the case studies in this part of the volume, taken
together, support a broad generalization. The damage potential of air
pollution, in economic terms, to commercial activities that use
biological systems (agriculture, forestry, and fisheries) does not
appear to be strikingly large. On the other hand, damages to
biologically-based recreational activities and to materials are
potentially very large but, at present, are also very ill understood.
94
-------
CHAPTER 15
BENEFITS FROM AVOIDING GROUNDWATER CONTAMINATION
While the extent of groundwater contamination is not accurately
known, the problem is thought to be widespread and is the focus of much
public apprehension. Contaminants in groundwater range across an
enormous list Of chemical substances, and usually no thorough checks for
contamination are made until there is reason to suspect a problem.
Even at extremely low concentrations, many toxic chemicals pose
serious, irreversible, health risks. In many of the cases checked, well
water has been found to contain concentrations above, and often several
orders of magnitude higher than, those commonly encountered in raw or
treated drinking water drawn from contaminated surface sources.
Thus, while water from most wells is no doubt safe, the widespread
nature of the contamination and its potential seriousness merit the
public attention the problem is getting. The intent of the case study
in this chapter is to develop methods for estimating benefits from
preventing contamination of groundwater-based drinking water supplies.
This, so far as I know, is the first study to attempt to quantify such
benefits. As in the studies discussed in other chapters, the
quantitative results reported here must be regarded as largely
experimental, but the numbers turn out to be impressively large.
For any chemical source, the extent of groundwater contamination
is determined by the characteristics of the underground storage medium--
called an aquifer. Groundwater in shallow, alluvial aquifers typically
moves less than a foot per day. That flow is governed by recharge and
discharge rates from the aquifer, and by the aquifer's permeability.
Contaminants are transported by diffusion together with the slow
underground flow of groundwater. In that oxygen-poor environment,
chemical or physical processes of contaminant degradation proceed very
slowly. Thus the contaminant plume may move great distances, with
hardly a change in toxicity levels, and may therefore reach drinking
water wells.
Among the principal sources of groundwater contamination are waste
disposal landfills and impoundments, accidental spills of chemical
substances, and abandoned oil and gas wells. Most groundwater
contamination can be traced to chemicals leaching into the aquifer from
poorly constructed and managed industrial or municipal landfills,
surface impoundments, or outright illegal dumps. Contamination from
such sources has often been in process for years, and sometimes for
decades. To date, most groundwater contamination incidents have been
discovered only after a
95
-------
drinking water source has been affected. By the time suspected aquifer
contamination is verified in samples drawn from drinking water wells,
the problem may be irreversible. Stricter regulation of the disposal of
potential contaminants in other environmental media, particularly air
and surface waters, and the consequent rising cost of such disposal, is
likely to increase the flow of wastes to land disposal and aggravate the
threat to groundwater.
Benefit analysis of controlling groundwater contamination
requires, as usual, quantification of several linkages between sources
and receptors. One must know the location and strength of actual or
potential sources of contamination. One must be able to model the
spread of the contaminant plume in the aquifer. One must know the
numbers of persons exposed to contaminated groundwater and the extent
and timing of their exposures. One must know the "dose-response
relationship," the nature and extent of health effects on the population
at risk. And finally, one needs a way of converting health effects into
monetary, or dollar values.
This is a very tall order, and we are far from being able to
quantify these linkages with precision. In each case, there is a need
for substantially improved methods and data. With these cautions in
mind, let us proceed to the case study. It involves the situation
associated with Price's landfill near Atlantic City, New Jersey.
Actually, while it is referred to as a landfill, this is rather
euphemistic--dump would be a better word, but I shall stick with the
conventional usage. Price's landfill occupies approximately twenty-two
acres extending across the boundary of Egg Harbor Township and the town
of Pleasantville, New Jersey. Until 1967, it functioned as a sand and
gravel quarry. During 1968, when the pit was excavated to within
approximately two feet of the water table, people from the surrounding
area began to dump trash into it with the permission of the owner,
Charles Price. In 1969, Price began commercial operations which
continued until the landfill was closed in 1976.
In 1970, Price applied to the New Jersey Department of
Environmental Protection for a license to conduct a sanitary landfill
operation. The application listed the materials that Price intended to
accept at the landfill, and specifically excluded "Chemicals (Liquid or
Solid)." He was issued a certificate authorizing operation of a solid
waste disposal facility.
In July 1972, authorities inspected the landfill, citing Price for
accepting chemical wastes and formally advising him of the violation.
Nonetheless, Price continued accepting significant quantities of
chemical wastes until November 1972. After that date, no chemical
wastes were disposed of at the landfill, although it continued in
operation. In 1976, Price terminated the landfill operation and covered
the site with fill material. The site has not been used since then.
But during the period May 1971 to November 1972, Price accepted
approximately 9 million gallons of the toxic and flammable chemical and
96
-------
liquid wastes, either in drums or directly into the ground. These
included (to name just a few) acids (glycolic, nitric, and sulfuric) ,
caustics and spent caustic wastes, cesspool waste, chemical resins and
other waste chemicals, chloroform, and cleaning solvents.
Price's Landfill is situated over the Cohansey aquifer, the
principal source of Atlantic City's water supply, and the separation
between landfill and aquifer is a relatively permeable layer. Waste
from the landfill is free to leach into the aquifer; the direction of
flow in the aquifer is eastward, which is toward Atlantic City's wells.
Chemicals in the leachate can therefore be carried into the private and
public water supply wells, and people can be exposed to those chemicals
in drinking water. Test wells drilled near the landfill by EPA show
that groundwater in the aquifer is contaminated and that the plume of
contamination is indeed moving toward Atlantic City's wells.
But estimation of actual or potential human exposures requires
either considerable information on, or heroic assumptions about, the
mechanism by which toxins are transported from the source of
contamination to the water supply wells. This is the second linkage
mentioned earlier. It will be clear shortly why discussion of this
linkage logically precedes the first quantification of the source of the
contamination.
Efforts to understand and model the source to receptor links,
called groundwater solute transport, are relatively recent. While there
has been considerable earlier work on salinity transport, study of the
more difficult cases of chemically reactive toxic groundwater
contaminants is less advanced. Improvements in our ability to model
these phenomena must be a prime objective for future research.
For purposes of analyzing the Price's Landfill situation, the
researchers chose and estimated numerically a technique called the
Wilson-Miller solute transport model. This relatively simple model was
chosen because of limitations of time and funding for the research.
The model chosen does appear to fit the Price's Landfill situation
relatively well and was judged adequate for conducting this experiment.
Future research should check to see if more complex models yield
substantially different results.
But to apply any solute transport model, it is necessary to have
so-called source-term information: the amounts of materials entering
groundwater and their distribution over time. This is the first linkage
mentioned earlier. Much of the activity at Price's landfill was
illegal. It therefore seems unlikely, to say the least, that careful
records of what went into the pit were kept. Indeed there is no
information at all about the amounts of the large number of chemical
substances dumped there. Where such records exist, or if leaching rates
are known or can be calculated, deliveries of pollutants to the aquifer
can be estimated directly. In the Price landfill type of situation,
typical of many existing groundwater contamination situations, there is
only one way to estimate the quantity of the source. Since we have
information on what is already present in test wells drilled by EPA, the
solute transport model can be run "backwards," so
97
-------
to speak, and used to infer what the amount of the source had to be to
produce the existing groundwater concentrations. This is why,
logically, discussion of the transport model precedes discussion of the
source term.
The reader should be cautioned that this estimate, while
necessary, is based on many assumptions and involves great uncertainty.
Just to give one example, the procedure assumes that releases occur at a
constant rate. This may not be true for some pollutants, and "slugs"
may be released which cause transients of pollution in much higher
concentrations than would be predicted by the model.
But given the computed source term, the model can be run
"forward," so to speak, to compute concentrations, at any well drawing
on the aquifer-the production wells of Atlantic City, for example--and
for any time after some contaminant enters the aquifer. Those
concentrations and the times at which they are projected to occur were
computed for the wells from which the Atlantic City Municipal Water
Authority pumps its water. Assuming that no mitigating action is taken,
this provides the link that specifies the exposure of the population to
contamination from Price's landfill.
To take the next step, one must have dose-response information--
that is, the actual health risk stemming from the contamination.
To make this link, information published by EPA was used. There is a
section of the Clean Water Act that requires EPA to estimate excess
cancer risks for 129 chemicals called "Priority Pollutants." Many of
these "Priority Pollutants" are ones leaching from Price's landfill into
the Cohansey aquifer. Using this information, the probability of excess
mortality from cancer was estimated for the population of Atlantic City.
While this procedure is the best available based on existing
information, the reader should be aware that, for this purpose, the risk
factors provided by EPA are both incomplete and very uncertain. For
example, there are many pollutants that have been identified in
groundwater that are not on the EPA list, and extrapolations from animal
toxicity tests to human risks are quite uncertain. In addition, it is
assumed that each chemical risk is independent of each other chemical
risk so that risks can simply be added up across chemical categories.
It is well known that "synergism" can occur which make the combined
toxicity of two chemicals greater than the sum of the effects of each
one taken independently.
Again, with all these cautions in mind, I turn to the next, and
final step, monetary evaluation of damages. The value of risk the
researchers chose to use is a range that reflects the underlying
uncertainty and reasonably well spans the range of values discussed in
chapter 4. The range chosen was from one hundred thousand dollars to
one million dollars per death. These values were then multiplied by the
mortality numbers calculated in the risk analysis to get a total benefit
from averting the damage which would otherwise emanate from Price's
landfill. The range turns out to be from 180 million dollars to 1.8
billion dollars.
Those are large figures, and one must be clear about what they mean.
Say that, at some site like the Price's landfill site, there is a
comparable release of contaminants into a similar aquifer, and that the
release goes unnoticed for two decades. Then there will be human
exposures through drinking water, and incremental mortality risks f aced
by the exposed population over their remaining lifetimes. Valuing this
98
-------
incremental mortality risk produced the above numbers. At a site at
which groundwater contamination has already occurred, those figures
represent the damages that might be avoided by measures taken to prevent
future exposures, either by restricting access to, or by cleansing, the
aquifer. Needless to say, those figures are impressively large. But the
limited information there is indicates that the costs of cleansing
aquifers are always large and the cost of obtaining an alternate water
supply may be large. This analysis, shaky as the numbers necessarily
are, suggests that in the case of groundwater contamination affecting
drinking water supplies, prevention is the best cure.
99
-------
CHAPTER 16
CONCLUDING NOTES
It seems fair to claim that the research reported in this volume
marks a substantial step forward in our ability to address the issue of
benefits from environmental quality improvement or maintenance.
Methods have been developed or improved, new data have been collected,
some case studies have been provided, and some highly preliminary
estimates of national benefits from environmental improvement or
maintenance have been presented. Furthermore, some broad insights
have resulted from the work. While so far I have done my best to fairly
state, in nontechnical terms, the findings of my colleagues in this
enterprise as they interpreted them, the following generalizations and
interpretations about findings are strictly my own.
Firstly, while our national air quality standards are based upon
alleged health effects, in fact, it appears from the work reported here
that we know very little for sure about the health consequences of air
pollution. The team's work on both aggregate and microepidemiology is
consistent with air pollution as a source of acute effects on an
important scale. However, human evidence of chronic effects is tenuous
at best. This is certainly not to say there are none, but conclusive
demonstration of such effects, or lack thereof, still awaits improved
data and methods.
Secondly, while our air quality standards are, as said, mostly
founded on presumed health impacts, it appears, based on the limited
evidence our studies were able to develop, that other economic damages
from pollution may be fully as great or even much greater. Damage to
materials appears to be a very large cost of poor air quality but, so
far, it has defied accurate quantification. In the preservation of
values area, it appears that protecting visibility, especially in the
West, yields large benefits. In the East, preventing deterioration of
water course recreational values through acid deposition appears to
involve large benefits. But, again, we are, alas, some distance from a
complete and accurate quantification of these values.
Thirdly, the interviewing done in connection with the "Visibility
in the National Parks" study suggests that there may also be a large
category of benefits which we have termed "intrinsic." That is, people
may be willing to pay for clean areas, in some cases on a really
substantial scale, even if they do not benefit directly from their use.
This may result from a feeling of national pride in having a clean
environment, especially in areas of outstanding natural beauty or
unusual cultural importance. Establishing these values in an accurate
and complete manner is still a frontier area in benefits research.
100
-------
Fourthly, in the area of water quality a large scale simulation
study suggests that the additional benefits to recreational fresh water
fishing from marginal improvements in water quality resulting from
implementation of national policy are not impressively large. This
is because so much of the nation's fresh water is already fishable.
However, an experimental national survey suggests that the willingness
of the public to pay to improve and maintain the quality of the nation's
water is large--on the order of many billions of dollars per year. This
research also suggests that a large portion, perhaps half, of these
benefits are of the non-user, intrinsic variety. This further suggests
that, in addition to the value of this type people may attach to some
particularly treasured sites, they may also find a large intrinsic value
in achieving certain nationally declared goals such like "swimmable"
waters virtually everywhere in the country. A full-scale national water
quality survey now underway and designed by members of the research team
should shed much additional light on the matter of both user and
intrinsic benefits
Fifthly, methods have been developed to study the agricultural
benefits of controlling air pollution. These, in contrast to earlier
studies, take account of various economic adaptations and adjustments,
for example crop or variety switching and the elasticity of demand for
agricultural products. Early findings suggest that while damages in a
highly polluted specialty crop area such as Southern California may be
significant, the main source of benefits from reduced pollution could
come from major field crops like soybeans and wheat. This is because
the total value of production of these crops is so huge that even a
relatively small increase in yields is associated with large benefits.
Sixthly, the groundwater "episodes" study implies that the
benefits from protecting large concentrations of population, such as the
Atlantic City area, from the toxic pollution of groundwater used for
drinking are potentially very large. In most cases they should easily
outweigh the costs of preventative measures.
Finally, I would like to close with some observations of a general
methodological character. The methods pursued in the studies discussed
here can be divided into two broad classes--those based, however
indirectly, on observed human behavior and those based on asking
questions about hypothesized situations. The former are based on actual
actions like travel to recreation sites and house prices paid. The
attraction of the behavior-based methods is that they reflect responses
to real, not hypothetical, situations and therefore are based on real,
not hypothetical decisions. But these behavior-based methods have
equally real limitations. For one thing, they are not applicable to all
situations of interest in environmental benefits evaluation, for
example, protecting a beautiful large vista from Visual impairment.
Further, they are limited to user benefits, and some of the research
surveyed here has suggested that intrinsic benefits may be very
important in certain cases.
For these reasons resort is made to methods based on asking
questions contingent on certain hypothetical situations. These are the
contingent valuation methods of bidding games and other surveys.
Inevitably, doubts
101
-------
arise about the accuracy of such methods given the hypothetical nature
of the situations they examine.
On the one hand the research reported here tends to support the
view that careful questionnaire design can control previously identified
sources of bias (starting point, strategic, etc.), and the South Coast
and San Francisco experiments tend to support the view that bidding
games can provide reasonable indicators of benefits from hypothetical
improvements in air quality, at least in certain instances. One reason
may be that persons residing in the regions studied, especially the Los
Angeles area, have a very clear understanding of the situation they find
themselves in and have mentally processed much information about it and
have taken decisions based upon it.
Very recent and highly preliminary experiments with bidding games
have suggested that where this close familiarity with the situation
being studied is not the case, a source of bias may exist that could
have substantial implications for some bidding game results. The
visibility in the parks study is perhaps the prime candidate among those
discussed in this volume. Recall that one interesting result of the
study was that the reported willingness to pay of respondents did not
appear to diminish with distance, e.g., those surveyed in Chicago had
fully as high a willingness to pay to protect visibility at the Grand
Canyon in the initial survey as those who were questioned in Denver.
In one set of later experiments, based on such a small sample that the
results should not be regarded as anything but suggestive of hypotheses
for future research, further bidding games were conducted in those two
cities. In both cities, instead of being asked questions only about
willingness to pay for visibility in the national parks, respondents
were first asked about their willingness to pay for other, closer to
home, environmental public goods. When this was done in Chicago,
willingness to pay for visibility in the national parks dropped sharply
below the result found in the previous survey. In Denver this was not
the case, perhaps because the questions about visibility were less
hypothetical to those in Denver and therefore their answer better
thought out than was true of respondents in Chicago. In another set of
experiments, again because of limited resources conducted with a highly
inadequate sample, persons were asked first about their willingness to
pay for a national improvement in water quality. Another sample was
then asked about the same improvement in water quality plus an
improvement in air quality. The resulting willingness to pay for both
was about the same as the willingness to pay for water quality
improvement alone in the case of the first group.
These kinds of highly experimental results have lead members of
the research team to speculate that people may have "mental accounts,"
one of which may be for environmental improvement. If this is the case,
when they are asked about a hypothetical, but rather dramatic,
environmental improvement they may allocate everything in their
environmental account to it, neglecting alternative environmental
improvements which, if confronted with them, they would also regard as
valuable. An important further development in contingent valuation
techniques will be to devise methods to
102
-------
structure them so as to avoid the one issue at a time procedure that has
characterized most applications so far.
In conclusion, while I believe that the research reported here
represents a significant improvement in our understanding of
environmental quality economic values, much remains to be learned.
Total accuracy about a matter of this difficulty is an impossible dream,
but I believe that the work done so far demonstrates that steady
progress is feasible.
103
-------
BIBLIOGRAPHY
REPORTS TO EPA UPON WHICH THIS VOLUME IS BASED
A. Other Volumes in this Series
Volume II
Six Studies of Health Benefits from Air Pollution Control
Shaul Ben-David, Reza Pazand, Scott E. Atkinson, Thomas D. Crocker,
Ralph C. d'Arge, Shelby Gerking, William D. Schulze, Curt Anderson,
Robert Buechley, Maureen Cropper, Lawrence A. Thibodeau, and Larry S.
Eubanks
Volume III
Five Studies on Non-Market Valuation Techniques
David S. Brookshire, William D. Schulze, Ralph C. d'Arge, Thomas D.
Crocker, Shelby Gerking, and Mark A. Thayer
Volume IV
Measuring the Benefits of Air Quality Changes in the San Francisco Bay
Area
Edna Loehman, David Boldt, and Kathleen Chaikin
Volume V
Measuring Household Soiling Damages from Suspended Particulates: A
Methodological Inquiry
R.G. Cummings, H.S. Burness, and R.D. Norton
Volume VI
The Value of Air Pollution Damages to Agricultural Activities in
Southern California
Richard M. Adams, Thomas D. Crocker, Narongsakdi Thanavibulchai, and
Robert L Horst, Jr.
Volume VII
Methods Development for Assessing Acid Deposition Control Benefits
Thomas D. Crocker, John T. Tschirhart, Richard M. Adams, and Bruce
Forster
Volume VIII
The Benefits of Preserving Visibility in the National Parklands of the
Southwest
William D. Schulze, David S. Brookshire, Eric G. Walther, Karen Kelley,
Mark A. Thayer, Regan L Whitworth, Shaul Ben-David, William Malm, and
John Molenar
Volume IX
Evaluation of Decision Models for Environmental Management
John Sorrentino
Volume X
Executive S
David S. Brookshire, Thomas D. Crocker, Ralph C. d'Arge, William D.
Schulze, Shaul Ben-David, Ronald G. Cummings, Allen V. Kneese, and Edna
Loehman
104
-------
105
-------
B. Subsequent Reports
Brookshire, David S., Ralph C. d'Arge, William D. Schulze, and Mark A.
Thayer. 1979. Experiments in Valuing Non-Market Goods: A Case Study of
Alternative Benefit Measures of Air Pollution Control in the South Coast
Air Basin of Southern California ' vol. 2 of Methods Development for
Assessing Tradeoffs in Environmental Management, EPA-600/5-79-001b.
Brookshire, David S., Thomas D. Crocker, Ralph C. d'Arge, Shaul Ben-
David, Allen V. Kneese, and William D. Schulze. 1979. Executive
Summary, vol. 5 of Methods Development for Assessing Air Pollution
Control Benefits, EPA-600/5-79-001E
Crocker, Thomas D., William D. Schulze, Shaul Ben-David, and Allen V.
Kneese. 1979. Experiments with Economics of Air Pollution Epidemiology,
vol. 1 of Methods Development for Assessing Air Pollution Control
Benefits, EPA-600/5-79-001a
Cropper, Maureen L., William R. Porter, Burton J. Hansen, Robert A.
Jones, and John G. Riley. 1979. Studies on Partial Equilibrium
Approaches to Valuation of Environmental Ameni-Eies, vol. 4 of Methods
Development for Assessing Air Pollution Control Benefits, EPA-600/5-19-
OOld.
Kopp, Raymond J., William J. Vaughan, and Michael Hazilla. 1983.
"Agricultural Sector Benefits Analysis for ozone: Methods Evaluation and
Demonstration," Final Report (Research Triangle Park, N.C., U.S.
Environmental Protection Agency, Office of Air Quality Planning and
Standards).
Loehman, Edna, David Boldt, and Kathleen Chaikin. 1980. Study Design and
Property Value Study, vol. 1 of Measuring the Benefits of Air Quality
Improvements in the San Francisco Bay Area, EPA-230-07-83-009.
Mitchell, Robert Cameron, and Richard T. Carson. 1981. "An Experiment
in Willingness to Pay for Intrinsic Water Pollution Control Benefits,"
Report to the U.S. Environmental Protection Agency (Washington, D.C.,
Resources for the Future).
Portney, Paul R., and John Mullahey. 1983. "Ambient Ozone and Human
Health: An Epidemiological Analysis," vols. I and II, Draft Final
Report, EPA-68-02358.
Schulze, William D., Ronald G. Cummings, and David S. Brookshire.
Experimental Approaches for Valuing Environmental Commodities.
Sharefkin, Mark F. 1983. "Economic Benefits from Control of Major
Environmental Episodes," (Washington, D.C., U.S. Environmental
Protection Agency, Office of Research and Development).
Vaughan, William J., and Clifford S. Russell. 1982. Freshwater
Recreational Fishing: The National Benefits of Water Pollution Control
(Washington, D.C., Resources for the Future).
ADDITIONAL PUBLICATIONS FROM EPA GRANTS
106
-------
FOR IMPROVING METHODS OF ESTIMATING BENEFITS FROM AIR QUALITY
Crocker, T.D., and D.S. Brookshire. "The Advantages of Contingent
Valuation Methods for Benefit Analysis," Public Choice 36 (1981), 235-
252 .
Crocker, T.D., R.M. Adams, and N. Thanavibulchai, "An Economic
Assessment of Air Pollution Damages to Selected Annual Crops in Southern
California," Journal of Environmental Economics, and Management 9(1)
(1982), 42-58.
Crocker, T.D., and B.A. Forster, "Decision Problems in the Control of
Acid Precipitation: Nonconvexities and Irreversibilities," Journal of
the Air Pollution Control Association 31 (January 1981), 31-38.
Crocker, T.D., "Hours of Work, Labor Productivity, and Environmental
Conditions: A Case Study," The Review of Economics and Statistics 63
(August 1981), 361-368.
Crocker, T.D., and R.M. Adams, "Analytical Issues in Economic
Assessments of Vegetation Damages," in P.S. Teng and S.V. Krupa, eds. ,
Crop Loss Assessments, St. Paul: University of Minnesota (1981), 221-
235 .
Crocker, T.D., "Conventional Benefit-cost Analyses of Acid
Deposition Control are Likely to Be Misleading," in P.S. Gold, ed., Acid
Rain:. A Transiurisdictional Problem in Search of Solution, Buffalo:
SUNY-Buffalo Press (1982), 7691.
Crocker, T.D., and R.M. Adams, "Economically Relevant Response
Estimation and the Value of Information: The Case of Acid Rain," in
T.D. Crocker, ed. , Economic Perspectives on Acid Deposition Control.,
Boston: Butterworth Scientific Publishers (1984), 35-64.
Crocker, T.D., "Scientific Truths and Policy Truths in Acid Deposition
Research," in T.D. Crocker, ed. , Economic Perspectives on Acid
Deposition Control, Boston: Butterworth Scientific Publishers (1984),
65-80.
Cropper, M.@, and A.S. Arriga-Salinas, "Inter-City Wage Differentials
and the Value of Air Quality," Journal of Urban Economics 8 (September
1980), 236-254.
, M.L, "Measuring the Benefits from Reduced Morbidity," The
American Economic Review 71 (May 1981), 235-240.
d'Arge, R.C., D. Brookshire, W. Schulze, and M. Thayer, "An Evaluation
of the Validity of Survey Methods in Economics," in P. Graves, G.
Tolley, and N. Cohen, eds., Environmental Policy and Air Quality,
Boston: Ballinger, (1981).
d'Arge, R.C., R. Rowe, and D. Brookshire, "An Experiment on the Economic
Valuation of visibility," Journal of Environmental Economics and
Management 7 (March 1980) .
d'Arge, R.C., W. Schulze, and D. Brookshire, "Economic Valuation of the
Risks and Impacts of Energy Development," in E.L. Etnier and C.C.
Travis, Health Risk Associated with Energy Technologies, Boulder,
107
-------
Colorado: Westview Press for the American Association for the Advan
t of Science (1981).
d'Arge, R.C., D. Brookshire, W. Schulze, and M. Thayer, "Experiments in
the Economic Valuation of Air Quality," in V. Kerry Smith, ed., Advances
in Applied Microeconomics, Nepw York: John Wiley (1981).
d'Arge, R.C., W. Schulze, and D. Brookshire, "Valuing Environmental
Commodities: Some Recent Experiments," Land Economics 57 (2) (November
1981), 151172.
d'Arge, R.C., D. Brookshire M. Thayer, and W. Schulze, "Valuing Public
Goods: A Comparison of Survey and Hedonic Approaches," American Economic
Review 72(1) (1982), 165-177.
Schulze, W., et al., "The Economic Benefits of Preserving Visibility in
the National Parklands of the Southwest," National Resources Journal
23 (i) (January 1983), 149-173.
Schulze, W., and S. Gerking, "What Do We Know About Benefits of Reduced
Mortality from Air Pollution Control?" American Economic Review,
Proceedings, 71 (May 1981), 228-234.
Schulze, W., and A. Kneese, "Risk in Benefit-Cost Analysis," Journal of
Risk Analysis 1, no. 1 (1981).
108
-------
READINGS*
Public Goods, Externalities, and Consumer's Surplus
Bishop, Richard C. and Thomas A. Herberlein. 1979. "Measuring Values of
Extra-Market Goods: Are Indirect Measures Biased?" American Journal of
Agriculture Economics vol. 61 (December), pp. 926-930.
Bohm, Peter. 1970. "Pollution, Purification, and the Theory of External
Effects," Swedish Journal of Economics, vol. 72, pp. 153-166.
Bohm, Peter. 1972. "Estimating Demand for Public Goods: An Experiment,"
European Economic Review, vol. 3, no. 2, pp. 111-130.
Bohm, Peter. 1979. '!Estimating Willingness to Pay: Why and How?"
Scandinavian Journal of Economics, vol. 81, m. 2.
Bowen, Howard R. 1943. "The Interpretation of Voting in the Allocation
of Economic Resources," Quarterly Journal of Economics, vol. 58, pp. 27-
48 .
Brookshire, David S., Ralph C. d'Arge, William D. Schulze, and Mark
Thayer. 1981. "Experiments in Valuing Public Goods," in V. Kerry Smith,
ed., Advances in Applied Microeconomics (Greenwich, Conn., JAI Press).
Ciccheti, Charles J., Anthony C. Fisher, and V. Kerry Smith. 1976. "An
Econometric Valuation of a Generalized Consumer Surplus Measure: The
Mineral King Controversy," Econometrica (November), pp. 1259-1276.
Gordon, Irene M., and Jack L. Knetsch. 1979. "Consumer's Surplus
Measures and the Evaluation of Resources," land Economics, vol. 55, no.
1 (February), pp. 1-10.
Harden, Russell. 1982. Collective Action (Baltimore, Md., Johns Hopkins
University Press for Resources for the Future).
Just, R.E., D.L. Hueth, and A. Schmitz. 1982. Applied Welfare Economics
and Public Policy (Englewood Cliffs, N.J., Prentice-Hall).
Kneese, Allen V., and Charles L. Schultze. 1975. Pollution., Prices,
and Public Policy (Washington, D.C., Brookings Institution).
Laffont, Jean-Jacques, ed. 1977. Aggregation and Revelation of
Preferences (Amsterdam, North-Holland)Lin, Steven, ed. 1976. Theory
and Measurement of Economic Externalities (New York, Academic Press)
Maler, Karl-Goran, and Ronald E. Wyzga. 1976. Economic Measurement of
Environmental Damage (Paris, Organization for Economic Cooperation and
Development) .
*This section was taken from Allen V. Kneese, Measuring the Benefits of
Clean Air and Water, 1984 (Washington ,D.C.: Resources for the Future,
Inc. )
109
-------
Mulligan, P.J. 1977. 'Willingness-to-Pay for Decreased Risk from
Nuclear Plant Accidents," Working Paper No. 3 (University Park,
Pennsylvania State University Energy Extension Program).
Pendse, Delip, and J.B. Whykoff. 1974. "Scope for Valuation of
Environmental Goods," Land Economics, vol. 50, no. 1 (February), pp. 89-
92 .
Scherr, Bruce A., and Emerson M. Babb. 1975. "Pricing Public Goods: An
Experiment with Two Proposed pricing Systems," Public Choice, vol. 23
(Fall), pp. 35-48.
Sinden, J.A., and A.C. Worrell. 1979. Unpriced Values: Decisions
Without Market Prices (New York, John Wiley).
Willig, R.D. 1976. "Consumer's Surplus Without Apology," The American
Economic Review, vol. 6, no. 4 (September), pp. 589-597.
Acid Deposition
"Acid Rain Research-A Special Report." 1983. EPRI Journal (Nov
Andersson, F. 1980. "Swedish Research on the Effects of Acid Deposition
on Forests and Water." Paper presented at the International Conference
on the Ecological Effects of Acid Precipitation, Oslo, Norway, March 11-
14 .
Appalachian Regional Commission. 1969. Acid Mine Drainage in Appalachia
(Washington, D.C.).
Bengtsson, B. 1980. "Liming Acid lakes in Sweden," Ambio, vol. 9, pp.
34-36 .
Blake, L.M. n.d. Liming Acid Ponds in New York (Watertown, New York
State Department of Environmental Conservation).
Carter, Luther J. 1979. "Uncontrolled C02 Emissions Bring Acid Rain,"
Science, vol. 204 (June 15), pp. 1181-1182.
Cogbill, C.V. 1976. "The History and Character of Acid Precipitation in
North America," Water, Soil, and Air Pollution, vol. 6, pp. 4 07-413.
Comptroller General of the United States. 1981. The Debate Over Acid
Precipitation: Opposing Views, Status of Research (Washington, D.C.,
Government Printing Office).
Dochinger, LS. , and T.X Seliga, eds. 1976. Proceedings of the First
International Symposium on Acid Precipitation and the Forrest Ecosystem
(Columbus, Ohio State University Press).
Galloway, J.N., and others. 1978. A National Program for Assessing the
Problem of Atmospheric Deposition (Acid Rain) Report to the President's
Council on Environmental Quality (Fort Collins, Colorado State
University, National Atmospheric Deposition Program).
Gold, Peter S., ed. 1982. Acid Rain: A Transiurisdictional Problem in
Search of Solution (Buffalo, Canadian-American Center, State University
of New
110
-------
Ill
-------
York)
Hendry, G., ed. 1978. Limological Aspects of Acid Precipitation,
BNL51074 Upton, N.J., Brookhaven National Laboratory).
Jacobson, J.S. 1981. "Acid Rain and Environmental Policy," Journal of
the Air Pollution Control Association, vol. 31, pp. 1071-1093.
Johnson, D.W. 1981. Acid Rain and Forest Productivity, Pub. No. 1717
(Oak Ridge, Tenn., Oak Ridge National Laboratory).
Jonsson, B., and R. Sundberg. 1972. "Has the Acidification by
Atmospheric Pollution Caused a Growth Reduction in Swedish Forests?"
Res. Note No. 20 (Stockholm, Sweden, Department of Forest Yield
Research, Royal College of Forestry) .
Northeast Regional Task Force on Atmospheric Deposition. 1981.
Northeast Damage Report on the Long-Range Transport of Air Pollutants
(Boston, Mass.).
Patrick, R.V., P. Binetti, and S.G. Halterman. 1981. "Acid lakes from
Natural and Anthropogenic Causes," Science, vol. 211 (January 30), pp.
446-448.
Wood, M.J. 1979. Ecological Effects of Acid Precipitation, Report of a
workshop held at Gatehouse-of-Fleet, United Kingdom, September 4-7, 1978
(Surrey, U.Y., Central Electricity Research Laboratories).
Air Pollution
Anderson, Robert C., and Bart Ostro. 1983. "Benefits Analysis and Air
Quality Standards," National Resources Journal, vol. 23, no. 3 (July),
pp. 56557 6.
Anderson, Robert J., Jr., and Thomas D. 1971. "The Economics of Air
Pollution: A literature Assessment," in P.B. Downing, ed., Air Pollution
and the Social Sciences (New York, Praeger), pp. 133-136.
Atkinson, S.E., and D.H. Lewis. 1974. "A Cost-Effectiveness Analysis of
Alternative Air Quality Control Strategies," Journal of Environmental
Economics and Management, vol. 1, pp. 2 37-2 50.
Atkinson, S.E., and D.H. Lewis. 1976. "Determination and Implementation
of Optimal Air Quality Standards," Journal of Environmental Economics
and Management, vol. 3 (September), pp. 363-380.
Barnes, RA 1979. "The Long Range Transport of Air Pollution: A Review
of the European Experience," Journal of the Air Pollution Control
Association, vol. 29 (December), pp. 1219-1235.
Barnett, L.B., and T.E. Waddell. 1973. The Cost of Air Pollution
Damages, Pub. No. AP-85 (Research Triangle Park, N.C. ', U.S.
Environmental Protection Agency) .
Freeman, A. Myrick III. 1978. "Air and Water Pollution Policy," in Paul
R. Portney, ed., Current Issues in U.S. Environmental Policy
(Baltimore, Md.,
112
-------
113
-------
John Hopkins University Press for Resources for the Future).
Heintz, H.T., Jr., A. Hershaft, and G.C. Horok. 1976. National Damages
of Air and Water Pollution, Report prepared for the U.S. Environmental
Protection Agency (Rockville, Md., Enviro Control, Inc.).
McDougall, Gerald S., and Colin Wright. 1980. "A Proposal for Improving
the Benefits from Pollution Abatement," Journal of Environmental
Economics and Management, vol. 7, no. 1 (March), pp. 0-29.
Mills, E.S., and L. White. 1978. "Auto Emissions: Why Regulation Hasn't
Worked," Technology Review (April-May).
National Academy of Sciences. 1974. Report by Committee on Motor
Vehicle Emissions (Washington, D.C., NAS).
National Academy of Sciences, Commission on Natural Resources. 1975.
Air Quality and Stationary Source Emission Control, U.S. Senate
Committee on Public Works, 94 Cong., 1st sess., Ser. 94-4 (Washington,
D.C., Government Printing Office).
National Academy of Sciences and National Academy of Engineering. 1974.
Quality and Automobile Emissions Control, vol. 4, The Costs and Benefits
of Automobile Emissions Control (Washington, D.C., NAS).
National Commission on Air Quality. 1981. To Breath Clean Air, Report
of the National Commission on Air Quality (W D.C., Government
Printing Office.)
Ridker, Ronald G. 1967. Economic Costs of Air Pollution (New York,
Praeger).
Rowe, Robert D., and Lorraine G. Chestnut, eds. 1983. Managing Air
Quality and Scenic Resources at National Parks and Wilderness Areas
(Boulder, Colo., Westview Press).
Ryan, John W. 1981. An Estimate of Nonhealth Benefits of Meeting the
Secondary National Ambient Air Quality Standards (Palo Alto, Calif.,
Stanford Research Institute, Inc.)
Seinfeld, John I-L 1975. Air Pollution: Physical and Chemical
Fundamenta1s (New York, McGraw-Hill) .
Smith, V. Kerry. 1977. The Economic Consequences of Air Pollution (Cam-
bridge, Mass., Ballinger).
Stern, A.C., ed. 1977. Air Pollution (3 ed., New York, Academic Press).
Waddell, T.E. 1974. The Economic Damages of Air Pollution,
Socioeconomic Environmental Studies Ties (Washington, D.C., U.S.
Environmental Protection Agency) .
White, Lawrence J. 1982. The Regulation of Air Pollutants Emissions
from Motor Vehicles (Washington D.C., American Enterprise Institute for
Public Policy Research).
114
-------
Benefit - Cost Analysis
Babcock, Lyndon R., and Miren L. Nagda. 1973. "Cost-Effectiveness of
Emission Control," Journal of the Air Pollution Control Association,
vol. 23, no. 3 (March), pp. 173-179.
Barnes, David W. 1983. "Back-door Cost-Benefit Analysis Under a Safety-
First Clean Air Act," Natural Resources Journal, vol. 23, no. 4
(October), pp. 827858.
Baumol, W.J., and W.E. Oates. 1979. Economics, Environmental Policy,
and the Quality of Life (Englewood Cliffs, N.J., Prentice-Hall).
Bradford, David F. 1970. "Benefit-Cost Analysis and the Demand for
Public Goods," Kyklos, vol. 23, p. 77 5.
Buchanan, J. 1969. Cost and Choice (Chicago, 111., Markham).
Burness, H. Stuart, Ronald G. Cummings, A.F. Mehr, and M.S.
Walbert. 1983. "Valuing Policies Which Reduce Environmental Risk,"
Natural Resources Journal 1, vol. 23, no. 3 (July), pp. 675-682.
Carson, Richard T., and Robert Cameron Mitchell. 1983. "Observations on
Strategic Bias and Contingent Valuation Surveys" (Washington, D.C.,
Resources
Costs and Benefits of Environmental Protection. 1981. (Canberra,
Australian Government Publishing Service)
Courant, Paul N., and Richard Porter. 1981. "Averting Expenditure and
the Cost of Pollution," Journal of Environmental Economics and
Management , vol. 8, pp. 321-329.
Freeman A. Myrick III. 1982. Air and Water Pollution Control: A
Benefit-Cost Assessment (New York, John Wiley).
Freeman, A. Myrick III. 1979. "The Benefits of Air and Water Pollution
Control: A Review and Synthesis of Recent Estimates," Report prepared
for the Council on Environmental Quality (Washington, D.C., CEQ).
Freeman, A. Myrick III. 1979. The Benefits of Environmental Improvement
(Baltimore, Md., John Hopkins University Press for Resources for the
Future)
Gramlich, Edward M. 1981. Benefit-Cost Analysis of Government Programs
Halvorsen, R., and M.G. Ruby 1983. Benefit Analyses of Air Pollution Control
(Lexington, Mass., Lexington Books).
Mathtech, Inc. 1982. Benefits Analysis of Alternative Secondary
National Ambient Air Quality Standards for Sulfur Dioxide and Total
Suspended Particulates (Princeton, N.J., Mathtech, Inc.)
Mills, E.S., ed. 1975. Economic Analysis and Environmental
Problems (New York, Columbia University Press for NBER).
National Academy of Sciences, Coordinating Committee on Air Quality
Studies.
115
-------
1974. The Costs and Benefits of Automobile Emission Control, vol. 4 of
Air Quality and Automobile Emission Control, Series No. 19-27
(Washington, D.C., NAS).
M., and Eugene P. Seskin, eds. 1975. Cost-Benefit Analysis of Water
Pollution Policy (Washington, D.C., Urban Institute.)
Randall, Alan, John P. Hoehn, and David S. Brookshire. 1983.
"Contingent Valuation Surveys for Evaluating Environmental Assets,"
Natural Resources Journal, vol. 23, no. 3 (July).
Ruff, L.E. 1970. "The Economic Common Sense of Pollution," The Public
Interest, vol. 19, pp.1 69-85.
Runge, Carlisle Ford. 1983. "Risk Assessment and Environmental Benefits
Analysis," Natural Resources Journal, vol. 23, no. 3 (July), pp. 683-
696 .
Schwing, R.C., W.S. Bradford, C.R. Von Busek, and C.J. Jackson. 1980. "Benefit-Cc
Analysis of Automotive Emission Reductions," Journal of Environmental Economics an
Management, vol. 7, no. 1 (March).
Swartzman, Daniel, Richard Leroff, and Kevin Croke, eds. 1982. Cost-Benefit
Analysis and Environmental Regulations (Washington, D.C., Conservation Foundation)
Water Resources Council. 1983. Economic and Environmental Principles
and Guidelines for Water and Related Land Resources Implementation
Studies (Washington, D.C., Government Printing Office).
Effect of Pollution on Agriculture
Adams, Richard M., and Thomas D. Crocker. 1982. "Dose-Response
Information and Environmental Damage Assessment: An Economic Perspective
," Journal of the Air Pollution Control Association, vol. 32, pp. 1062-
1067 .
Benedict, H.M., C.J. Miller, and J.S. Smith. 1973. Assessment of
Economic Impact of Air Pollutants on Vegetation in the United States--
196 9 and 1971 (Menlo Park, California., Stanford Research Institute).
Carriere, W.M., A.D. Hinkley, W. Harshbarger, J. Kinsman, and 7.
Wisniewski. 1982. "The Effect of S02 and 03 on Selected Agricultural
Crops," Report prepared for the Electric Power Research Institute
(McLean, Va., General Research Corporation).
Crocker, Thomas D., B.@ Dixon, R.E. Howitt, and R. Oliveria. 1981. "A
Program for Assessing the Economic Benefits of Preventing Air Pollution
Damages to U.S. Agriculture," Discussion Paper prepared for the National
Crop Loss Assessment Network (NCLAN), (Laramie, University of Wyoming).
Heck, W.W., O.C. Taylor, R. Adams, G. Bingham, J. Miller, E. Preston,
and L. Weinstein. 1982. "Assessment of Crop Loss from Ozone," Journal
of the Air Pollution Control Association, vol. 32, no. 4, pp. 353-361.
Heck, W.W., O.C. Taylor, R. Adams, G. Bingham, J.E. Miller, and @H.
Weinstein. 1981. National Crop Loss Assessment Network 1980 Annual
Report (Corvallis, Ore., U.S. Environmental Protection Agency,
116
-------
Environmental Research laboratory, Office of Research and Development
) .
Jacobsen ,J.S. and A. A. Miller, eds. 1982. Effects of Air Pollution on
Farm Commodities (Arlington, Va., Izaak Walton League of America), pp.
103-124.
Johnston, W.E., and G.W. Dean. 1969. California Crop Trends: Yields,
Acreages, and Production Areas, California Agricultural Experiment
Station, Extension Service Circular 555 (Berkeley, University of
California).
Research Management Committee. 1982. The National Crop Loss Assessment
Network (NCLW: 1981 Annual Report (Corvallis, Ore., Environmental
Research Laboratory, U.S. Environmental Protection Agency).
Southern California Association of Governments and South Coast Air
Quality Management District. 1979. Air Quality Management Plan.
Thomas, M.D. 1961. "Effects of Air Pollution on Plants," in Air
Pollution, World Health Organization Monograph Series No. 46 (New York,
Columbia University Press), pp. 233-278.
Effect of Pollution an Ecosystems
Crocker, Thomas D., John T. Tschirhorst, and Richard M. Adams. 1980.
"Valuing Ecosystem Functions: The Effects of Acidification." Paper
prepared for Redistributive Impact in the Federal Acid Rain Program,
Denver, Colo., September, 5.
Fisher, Anthony C., and John V. Yxutilla. 1974. "Valuing
Long-run Ecological Consequences and Irreversibilities, "Journal of
Environmental Economics and Management.
Fisher, Anthony C., John V. Krutil la, and Charles J. Ciechetti. 1972.
"The Economics of Environmental Preservation: A Theoretical and
Empirical Analysis," American Economic Review.
Hagerhill, B. 197 9. "Estimation of Economic Damages to Aquatic
Ecosystems in Sweden in Relation to Different Emissions Control Cases
for 1985." Paper prepared for the Organization of European Co-Operation
and Development Workshop on Acid Precipitation, Paris, France, April 24-
26
Hirschleifer, Jack. 1977. "Economics
Journal of law and Economics, vol. 2 0
Hutchinson, T.C., and M. Havas,
Precipitation on ial Ecosystems. (New
from a Biological Standpoint," The
(April), pp. 1-52.
eds. 1980. Effects of Acid
York, Plenum Press).
117
-------
Kneese, Allen V. 1976. "Analysis of Environmental Pollution," Swedish
Journal of Economics, vol.3 pp. 253-288.
Kormondy, E.J. 1969. Concepts of Ecology (Englewood Cliffs, N.J.,
Prentice Hall
Tullock, Gordon. 1971. "Biological Externalities," Journal of
Theoretical Biology, vol. 33, pp. 565-576.
Woodwell, G.M. 1970. "Effects of Pollution on the Structure and
Physiology of Ecosystems," Science, vol.168 (April), pp.429-433
Effect of Pollution m Human Health
Arthur, W.B. 1981. "The Economics of Risks to Life," American Economic
Review, vol. 71, no. 1 (March).
Bailey, Martin J. 1980. Reducing Risks to Life: Measurement of the
Benefits (Washington, D.C., American Enterprise Institute).
Ehagia, Gcbend S., and Herbert Stoevener. 1978. 'Effect of Air
Pollution on Consumption of Medical Services (Corvallis, Ore., U.S.
Environmental Protection Agency).
Bouhuys, A., G.J. Beck, and J.B. Schoenberg. 1978. "Do Present levels
of Air Pollution Outdoors Affect Respiratory Health?" Nature, vol. 276
(Nov 30), pp. 466-471.
Carpenter, Ben H., D. A. LeSourd, James R. Chromy, and Walter D. Boch.
1977. Health Costs of Air Pollution Damages, A Study of Hospitalization
Costs (Research Triangle Park, N.C., U.S. Environmental Protection
Agency).
Christiansen, G., and C. Degan. 1980. "Air Pollution and Mortality
Rates: A Note on Lave and Seskin's Pooling of Gross-Section and Time-
Series Data," Journal of Environmental Economics and Management, vol. 7,
no. 2 (June), pp. 149-155.
Colmar, C.L, and L.A.Sagan. 1976. "Health Effects of Energy Production
and Conversion," in J.M. Hollander, ed., Annual Review of Energy, vol.
1, pp. 581600.
Freeman, A. Myrick III. 1982. "The Health Implications of Residuals
Discharges: A Methodological Overview," in V. Kerry Smith and John V.
Krutilla, eds. , Explorations in Natural Resource Economics (Baltimore,
Md., Johns Hopkins University Press for Resources for the Future).
Goldsmith, J.R., and L.T. Fribert. 1977. "Effects of Air Pollution on
Human Health," in A.C. Stern, ed.,, The Effects of Air Pollution (3 ed.,
New York, Academic Press).
Gotchy, R.L 1977. "Health Effects Attributable to Coal and Nuclear Fuel
Cycle Alternatives" (Washington D.C., U.S. Nuclear Regulatory
Commission).
118
-------
Higgins I.T.T. 1974. Epidemiology of Chronic Respiratory Disease: A
Literature Review, Environmental Health Effects Research Series, EPA-
650/1-74-007 (Research Triangle Park, NC, U.S. Environmental Protection
Agency).
Kneese, Allen V., and William Schulze. 1977. "Environment, Health, and
Economics--The Case of Cancer," American Economic Review, no. 67
(February), pp. 26-32.
Lave, Lester B., and Eugene P. Seskin. 1970. "Air Pollution and Human
Health," Science, vol. 169 (August), pp. 723-733.
Lave, Lester B. 1977. Air Pollution and Human Health (Baltimore, Md.,
Johns Hopkins University Press for Resources for the Future).
Lipfert, F. 1980. "Sulfur Oxides, Particulates and Human Mortality:
Synopsis of Statistical Correlation," Journal of the Air Pollution
Control Association, vol. 31.
National Academy of Sciences. Committee on Medical and Biological
Effects of Environmental Pollutants. 1977. Ozone and Other Photochemical
Oxidants (Washington, D.C., NAS).
Ostro, Bart D. 1983. "The Effects of Air Pollution on Work Loss and
Morbidity," Journal of Environmental Economics and Management, vol. 10,
no. 4 (December),, pp. 371-382.
Page, Talbot, Robert Harris, and Samuel S. Epstein. 1976. "Drinking
Water and Cancer Mortality in Louisiana," Science, vol. 193, no. 4247
(July), pp. 55-57.
Perara, Fredrick P., and A. Karim Ahmed. 1979. Respirable Particles:
Impact of Airborne Fine Particulates on Health and the Environment
(Cambridge, Mass., Ballinger), pp.31-44.
Seskin, Eugene P. 197 9. "An Analysis 0 Some Short-Term Health Effects of
Air Pollution in the Washington, D.C., Metropolitan Areas," Journal of
Urban Economics, vol. 6(July), pp. 275-291.
Thibodeau, L. 1980. "Air Pollution and Human Health: A Review and
Reanalysis," Environmental Health Perspectives, vol. 34.
U.S. Environmental Protection Agency. 1974. Health Consequences of
Sulfur Oxides: A Report from CHESS, 1970-71 (Research Triangle Park,
N.C., EPA Human Studies Laboratory).
Viren, John R. 1978. Cross-Sectional Estimates of Mortality Due to
Fossil Fuel Pollutants: A Case for Spurious Association (Washington,
D.C., U.S. Department of Energy).
Effect of Pollution on Materials
Booz-Allen and Hamilton, Incorporated. 1970. Study to Determine
Residential Soiling Costs of Particulate Air Pollution (Washington,
D.C. ) .
119
-------
Crocker, Thomas D., and Ronald G. Cummings. 1983. "On Valuing
Deposition-Induced Materials Damages: A Methodological Inquiry"
(Laramie, University of Wyoming).
Gillette, Donald D. 1975. "Sulfur Dioxide and Material Damage," Journal
of the Air Pollution Control Association, Vol. 25, no. 12 (December),
pp. 1238-1243.
Watson, William D., and John A. Jaksch. 1982. "Air Pollution: Household
Soiling and Consumer Welfare Losses," Journal of Environmental Economics
and Management , vol.9, no. 3 (September) pp.248-262.
Yocon, V.E., and N. Grappom 1976. "The Effects of Power Plant Emissions
on Materials," Report prepared by the Research Corporation of New
England (Palo Alto, Calif., Electric Power Research Institute).
Effect of Pollution on Property Values
Abelson, Peter W. 1979. "Property Prices and the Value of Amenities"
ties," Journal of Environmental Economics and Management, Vol. 6, no. 1
(March), pp. 110.
Anderson, Robert J., and Thomas D. Crocker. 197 1. "Air Pollution and
Residential Property Values," Urban Studies, No. 8 (October), pp. 171-
180 .
Brigham, Eugene F. 1965. "The Determinants of Residential Land
Values,," Review of Economics and Statistics, Vol. 59 (August), pp. 272-
278 .
Deyah, Timothy, and V. Kerry Smith. 1974. "Residential Property Values
and Air Pollution: Some New Evidence," Quarterly Review of Economics and
Business, Vol. 14 (Winter), pp. 93-100.
Freeman, A. Myrick III. 1971. "Air Pollution and Property Values: A
Methodological Comment," Review of Economics and Statistics, Vol. 53,
pp. 415-416.
Freeman, A. Myrick III. 1974. "On Estimating Air Pollution Control
Benefits from Land Value Studies," Journal of Environmental Economics
and Management, Vol. 1, no. 1 (May), pp. 74-83.
Harris, R.N.S., G.S. Tolley, and C. Harrell. 1968. "The Residence Site
Choice," Review of Economics and Statistics, Vol. 50 (May), pp. 241-247.
Harrison, David, Jr., and Daniel L Rubinfeld. 1978. "The Air Pollution
and Property Value Debate," Review of Economics and Statistics, Vol. 60,
no. 4 (November), pp.635-638.
Hoch, Irving. 1977. "Variations in the Quality of Urban Life Among
Cities and Regions," in Lowden Wingo and Alan Evans, eds., Public
Economics and University Press.
Harrison, David, Jr., and Daniel L. Rubinfeld. 1978. "Hedonic Housing
Prices and the Demand for Clean Air," Journal of Environmental Economics
and Management, vol. 5, no. 2 (March), pp. 81-102.
Hoch, Irving. 1977. "Variations in the Quality of Urban Life Among
Cities and Regions," in Lowden Wingo and Alan Evans, eds., Public
120
-------
(Baltimore, MD., Johns Hopkins University
Press for Resources for the Future), pp.28-65.
Nelson, J. 1979. "Airport Noise, Location Rent, and the Market for
Residential Amenities," Journal of Environmental Economics and
Management, vol. (December), pp.320-331.
Niskanen, W-k, and S. FL Hanke. 1977. "Land Prices Substantially
underestimate the Value of Environmental Quality," Review of Economics
and Statistics, vol. 59, pp. 375-377.
Polinsky, A.M., and S. Shavell. 1975. "The Air Pollution and Property
Value Debate," Review of Economics and Statistics, vol. 57, pp. 100-104.
Portney, Paul R. 1981. "Housing Prices, Health Effects, Valuing
Reductions in Risk to Death," Journal of Environmental Economics and
Management, vol.8.,pp.72-78.
Ridker, Ronald G., and John Heming. 1967. "The Determinants of
Residential Property Values With Special Reference to Air Pollution, "
Review of Economics and Statistics, vol. 49 (May), pp. 246-257.
Small, K. A. 1975. "Air Pollution and Property Values: A Further
Comment," Review of Economics and statistics, vol. 57, pp. 105-107.
Smith, V. Kerry, and Timothy A. Deyak. 1975. "Measuring the Impact of
Air Pollution on Property Values," Journal of Regional Science, vol. 15,
no. 3, pp. 277-288.
Tolley, George S., and Allen S. Cohen. 1976. "Air Pollution and Urban
Land Use Policy," Journal of Environmental Economics and Management,
vol. 2, pp. 247-254.
Wieand, K.F. 1973. "Air Pollution and Property Values: A Study of the
St. Louis Area", Journal of Regional Science, vol. 13, pp. 91-95.
Effects of Pollution on Visibility
Blank, F., David S. Brookshire, Thomas D. Crocker, Ralph C. d'Arge, R.
Horst, Jr., and R. Rowe. 1977. Valuation of Aesthetic Preferences: A
Case Study of the Economic Value of Visibility (Laramie, University of
Wyoming, Resource and Environmental Economics Laboratory).
Brookshire, David S., Berry C. Ives, and William D. Schulze. 1976. "The
Valuation of Aesthetic Preferences," Journal of Environmental Economics
and Management. vol. 3 , no. 4
Charles River Associates, Incorporated. 1980. "Visibility at mesa
Verde National Park: An Analysis of Benefits and Costs of Controlling
Emissions in the Four Corners Area," Report prepared for the Electric
Power Research Institute (Palo Alto, Calif., EPRI).
Charlson, R. 1978. "Seven Authorities Speak Out on Visibility," Report
prepared for the National Park Service, U.S. Department of the Interior
(Washington, D.C., Government Printing Office).
Rae, Douglas A. 1982. "The Value to Visitors of improving Visibility at
Mesa Verde and the Great Smoky National Parks," in Robert D. Rowe and
121
-------
Lauraine G. , eds., Managing Air Quality and Visual Resources at
National Parks and Wilderness Areas(Boulder, Colo., Westview Press).
Randall, Alan, Berry Ives, and Clyde Eastman. 1974. Benefits of
Abating Aesthetic Environmental Damage from the Four Corners Power
Plant, Fruitland. New Mexico, Bull. 618 (Las Cruces, New Mexico State
University Agriculture Experiment Station).
Randall, Alan, Berry Ives, and Clyde Eastman. 1974. "Bidding Games for
Evaluation of Aesthetic Environmental Improvement," Journal of
Environmental Economics and Management, vol. 1, no. 2 (August), pp. 132-
149 .
Rowe, Robert D., and Lauraine G. Chestnut. 1982. The Value of
Visibility . Economic Theory and Applications for Air Pollution
Cambridge, Mass., ABT Books).
Trijonis, J., and K. Yuan. 1978.
(Research Triangle Park, N.C., U.S.
Office of Research and Development,
Laboratory).
"Visibility in the Northeast"
Environmental Protection Agency,
Environmental Sciences Research
Trijonis, J., and Y- Yuan. 1978. "Visibility in the Southwest: An
Exploration of the Historical Data Base" (Research Triangle Park, N.C.,
U.S. Environmental Protection Agency, Office of Research and
Development, Environmental Sciences Research Laboratory).
Outdoor Recreation
Battelle Memorial Institute. 1975. "Assessment of the Economic and
Social Implications of Water Quality Improvements on Public Swimming"
(Columbus, Ohio).
Bishop, Richard C., and Thomas A. Heberlein. 1980. "Simulated Markets,
Hypothetical Markets, and Travel Cost Analysis: Alternative Methods of
Estimating Outdoor Recreation Demand," Wisconsin Agriculture
Experimental Station Bulletin, no. 187.
Burt, Oscar R., and Durwood Brewer. 1971. "Estimation of Net Social
Benefits from Outdoor Recreation," Econometrica, vol. 39, no. 5
(September), pp. 813-827.
Cicchetti, Charles J. 1973. Forecasting Recreation in the United States
(Lexington, Mass., Lexington Books).
Cicchetti, Charles J., and V. Kerry Smith. 1976. The Cost of Congestion:
An Econometric Analysis of Wilderness Recreation (Cambridge, Mass.,
Ballinger).
Clawson, Marion. 1959. "Methods of Measuring Demand for and Value of
Outdoor Recreation," Resources for the Future Reprint No. 10
(Washington, D.C., Resources for the Future).
Clawson, Marion, and Carlton S. Van Doren, eds. 1984. Statistics on
Outdoor Recreation, Part I, The Record Through 1956; Part II, The Record
Since 1956 (Washington, D.C., Resources for the Future).
122
-------
Desvouges, William H., V. Kerry Smith, and Matthew P. McGivney. 1983. A
Comparison of Alternative Approaches for Estimating Recreation and
Related Benefits of Water Quality Improvements (Research
Triangle Park, N.C., Research Triangle Institute).
Greenley, Douglas, Richard G. Walsh, and Robert A. Young. 1981. "Option
Value: Empirical Evidence from a Case Study of Recreation and Water
Quality," The Quarterly Journal of Economics, vol. 96 (November), pp.
657-673.
Krutilla, John V. 1975. The Economics of Natural Environments: Studies
in the Valuation Commodity and Amenity Resources
(Baltimore, Md. , Johns Hopkins University Press for Resources for the
Future).
McConnell, Kenneth. 1975. "Some Problems in Estimating the Demand
for outdoor Recreation," American Journal of Agricultural Economics,
vol. 57, pp. 330-334.
Majiid, I., J.X Sinden, and Alan Randall. 1983. "Abstract: Benefit
Evaluation of Increments to Existing Systems of Public Facilities," Land
Economics (November).
Mendelsohn, Robert, and Gardner M. Brown, Jr. 1983. "Revealed
Preference Approaches to Valuing Outdoor Recreation, " Natural Resources
Journal, vol. 23, no. 3 (July), pp. 607-618.
Moeller, George FL, and John IL EnglekexL 1972. "What Fishermen Look
for in a Fishing Experience," Journal of Wildlife Management, vol. 36,
no. 4 (October).
Rankin, R.L., and J.A. Sinden. 1971. "Causal Factors in the Demand for
Outdoor Recreation," Economic Record, vol. 47 (September), pp. 418-426.
Russell, Clifford S., and William J. Vaughan. 1982. "The National
Recreational Fishing Benefits of Water Pollution Control," Journal of
Environmental Economics and Management, vol. 9, no. 4 (December), pp.
328-354 .
Stevens,JoeB 1966. "Recreation Benefits from Water Pollution
Control," Water Resources Research, vol. 2, no. 2, pp. 167-182.
Sutherland, Ronald J. 1982. "A Regional Approach to Estimating
Recreation Benefits of Improved Water Quality," vol. 9, no. 3
pp. 229-247.
Water Pollution
Bouwes, Nicholas W. , and Robert Schneider. 1979. "Procedures in
Estimating Benefits of Groundwater Quality Change," American Journal of
Agricultural Economics, vol.61, no.3 (August), pp.535-539.
Council on Environmental Quality. 1981. "Contamination of Groundwater
by Toxic Organic Chemicals" (Washington, D.C., CEQ).
Cronin, Francis J. 1981. "A Contingent Market Valuation of Improved
Water Quality," mimeo. (Richmond,Wash., Battelle Pacific Northwest
Laboratory).
123
-------
Daniel, and Edwin S. Mills. 1980. Measuring the Benefits of Water
Pollution Abatement (New York, Academic Press).
Greenley, Douglas A., Richard G. Walsh, and Robert A. Young,
eds. 1982. Economic Benefits of Improved Water Quality: Public
Perception of Option and Preservation Values (Boulder, CO., Westview
Press) .
Hanemann, W.M. 1978. A Methodological and Empirical Study of the
Recreation Benefits from Water Quality Improvement (Berkeley, University
of California, Department of Agriculture and Resource Economics).
Unger, S.G. 1976. National Benefits of Achieving the 1977, 1983, and
1985 Water Quality Goal., Final Report for the U.S. Environmental
Protection Agency (Washington, D.C., Development Planning and Research
Associates, Inc.).
General Resource Economics
P.S., and G.M. Heal. 1979. Economic Theory and Exhaustible Resources
(New York, Cambridge University Press).
Fisher, Anthony C. 1981. Resource and Environmental Economics
(New York, Cambridge University Press).
Herfindahl, Orris C., and Allen V. Kneese. 1974. Economic Theory of
Natural Resources (Columbus, Ohio, Charles E. Merrill).
Howe, Charles W. 1979. Natural Economics (New York, John Wiley).
Maler, Karl-Goran. 1974. Environmental Economics: A Theoretical Inquiry
(Baltimore, Md., Johns Hopkins University Press for Resources for the
Future).
Smith, V. Kerry, and John V. Krutilla, eds. 1982. Explorations in
Natural Resource Economics (Washington, D.C., Resources for the Future).
Tietenberg, Thomas H. 1984. Environmental and Natural Resource
Economics: Analysis and Policy (Glenview, 111., Scott-Foresman).
124
------- |