vvEPA
United States
Environmental Protection
Agency
Office Of Water
(WH-556)
EPA800-R-93-001b
March 1993
Clean Water And The
American Econnomy
Proceedings: Ground Water
Volume 2
October 19-21,1992
Recycled/Recyclable
Printed on paper that contains
at least 50% recycled fiber
-------
DISCLAIMER
The mention of commercial products, their source, or their use in connection with
material reported herein is not to be construed as either an actual or implied
endorsement of such products by the U.S. Environmental Protection Agency. The
content of all papers is as presented at the conference. Audio transcriptions are
presented verbatim with no editorial changes. Format and grammatical changes have
been made where possible for consistency.
U S Environmental Protection Agency
Region 5, Library (PL-12J)
77 West Jackson Boulevard, 12th Floor
Chicago, IL 60604-3590
-------
ACKNOWLEDGEMENTS
The U.S. Environmental Protection Agency wishes to acknowledge the following
organizations for their support and help in conducting this conference. Resources for the
Future co-sponsored the conference. The Ground Water Protection Council provided
funding and assistance in scheduling speakers. This document was prepared under the
direction of Mark Luttner, Special Assistant to the Assistant Administrator, Office of
Water and Charles Job, Chief of the Ground Water Protection Branch. The following
Radian Corporation personnel provided logistical and technical support in the
development of this document for the Office of Water, under EPA Contract
No. 68-CO-0032: Thomas Grome, Program Manager; William Sproat, Project Director;
Margaret Masley, Task Leader; Laurie Morgan, Chrisanti Haretos, Margie Gibson, and
Robyn Flanders.
-------
22 VOLUME II
TABLE OF CONTENTS
Page
James R. Elder-Ground Water Valuation Overview GW1-2
V. Kerry Smith- Valuing Ground Water Resources: A Conceptual Overview GW1-5
William Schulze-Do Existence Values Exist for Ground Water? Methods and Case Studies GW2-2
John C. Berg^tTora-Benefits of Protecting Ground Water from Agricultural Chemical Contamination . GW2-6
Charles W. Abdaftar-Avoidance Costs and Ground Water Values: Results of Two
Empirical Applications GW2-17
Erik Lichtenberg-l/H/ig Risk Assessment in Analyzing Ground Water Protection Policies GW2-24
R. Gregory Michaels- When the Home Is No Longer a Castle: Inferring the Economic Value of Ground Water
Contamination from Residential Property Sales GW2-34
Questions and Answers GW2-42
Velma Smiib-Policy and Regulatory Considerations GW3-2
James A. Goodrich- Ground Water Value in California GW3-5
William P. Weisrock— Ground Water Valuation: An Industry Perspective GW3-10
Paul Jehn- Ground Water Valuation: The High Cost of Not Protecting Ground Water GW3-19
Keith Cole-Policy and Regulatory Considerations GW3-26
Jimmie Powell-Policy and Regulatory Considerations GW3-29
Debra Jacobson-/>o//cy and Regulatory Considerations GW3-33
Questions and Answers GW3-36
Marjorie M. Holland-/y.yuey and Research Needs in Valuing Ground Water: An Ecosystem
Perspective GW4-2
Robert Costanza-Issues and Research in Valuing Ground Water GW4-6
Eric J. Harmon, P.E.—Liquid Assets and Paper Water: Valuation of Ground Water Under
Colorado's Prior-Appropriation System Traditional Basis and New Issues GW4-9
Stephen R. Crutchfield-/y,suey in Measurement of the Non-Use Value of Ground Water GW4-17
Maureen L. Cropper— Research Issues in Valuing Ground Water GW4-28
Richard B. Howarth-Environmental Risks and Future Generations: Criteria for Public Policy GW4-31
Questions and Answers GW4-43
-------
Session GW-1
Summary
Session GW-1: Ground Water Valuation Overview
SESSION SUMMARY
PRESENTERS:
James R. Elder-Ground Water Valuation Overview
V. Kerry Smith- Valuing Ground Water Resources: A Conceptual Overview
GW1-1
-------
Session GW-1
Elder
Ground Water Valuation Overview
James R. Elder
Director
Office of Ground Water and Drinking Water
U.S. Environmental Protection Agency
Alternative approaches and reasons for valuing ground water affect nearly every program
that the Environmental Protection Agency implements. However, there is inconsistency among EPA's
programs in determining ground water value. There is also inconsistency among our federal statutes. This
program will present the most recent information on approaches for valuing ground water, along with a
discussion of policy and regulatory issues. As we review the current work in ground water valuation, we can
identify issues needing attention by EPA. This information will be used by the EPA as it begins the debate
on establishing a common basis for valuing ground water across its programs.
Ground water is a vital resource for the United States. Fifty percent of the nation's
population relies on ground water drinking water supplies. In several states, such as Florida, that figure is
over 80%. Ninety-five percent of our rural population drink ground water, and the untapped potential is
enormous. Ninety-five percent of the Earth's non-frozen fresh water supplies are ground water.
Ground water has many uses other than as a drinking water supply. It is also used for
irrigation, industry, livestock, mining, and thermoelectric power. Sixty-two percent of ground water
withdrawals are used for irrigation. Twenty-four percent of withdrawals are used for domestic consumption.
Ground water supplies 68% of livestock needs and 59% of mining needs.
U.S. fresh ground water withdrawals average approximately 73 billion gallons per day. One-
fourth of this total occurs in two states: California and Texas. Other major users include Idaho, Kansas and
Nebraska. These withdrawals are primarily for irrigation.
From a historical standpoint, many locations in the United States would never have
developed and become as productive as they are if the ground water resource had not been of high quality
and readily accessible. For example, ground water has been vital to the farming, ranching, and mining efforts
in the High Plain States and much of the rest of the nation.
Ground water also plays a significant role in maintaining the ecosystem. In my office's
latest assessment of ground water/surface water interaction, we have documented over 100 different studies
addressing this interaction. These studies have shown the vital role ground water plays in maintaining our
surface water. Forty percent of average annual steam flow in the U.S. is base flow maintained by ground
water discharge. Under drought conditions this percentage increases. Ground water also constitutes a larger
percentage of stream flow in humid areas such as the northeast. In the Delmarva Peninsula, this ground
water discharge can be as much as 80% of the total stream flow. The health of many surface water
ecosystems, such as wetlands and estuaries, is dependent on the ground water/surface water interaction.
The increasing awareness of the importance of ground water ecosystems and how they affect
surface water ecosystems, induced EPA to co-sponsor the first international conference on Ground Water
Ecology held last spring in Tampa. At this conference, leading scientists from around the world spoke on
their latest discoveries—including the importance of ground water/surface water interaction and its role in
maintaining the health of rivers; microscopic and macroscopic organisms that live in some parts of ground
water; and the knowledge of how ground water ecology can be the basis of managing ground water for
ecological benefits. EPA is now developing a ground water ecology initiative as part of its larger Ground
GW1-2
-------
Session GW-1
Elder
Water Strategy that will not only address protection of ground water and surface water ecosystems, but will
help ensure cleaner and safer sources of ground and surface water for human use.
The importance of ground water has also been recognized by Congress and EPA. EPA's
statutes clearly assign a stewardship role to EPA for ground water quality and reflects society's decision to
protect the resource for the benefit of current and future uses. A few of the major statutes are: the
Resource Conservation Recovery Act (RCRA); the Comprehensive Environmental Response Compensation
and Liability Act (CERCLA); the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA); the Atomic
Energy Act; the Clean Water Act; and the Safe Drinking Water Act (SDWA). In response to these Acts
and others, EPA has 20 programs that address ground water protection, ranging from the well-head
protection and sole-source aquifer programs to the hazardous waste and Superfund programs.
Most of the statutes affecting ground water, including RCRA, CERCLA, SDWA and the
Clean Water Act are scheduled for reauthorization by Congress. Clearly, any decision that Congress makes
affecting ground water value will have profound implications on our programs, and through our programs,
the states and localities. For example, in the RCRA reauthorization process, the House Energy and
Commerce Committee added report language recognizing different beneficial uses of ground water in
Subtitle D. This allows for different values of ground water based on its use. Currently, all our legislation is
silent on this issue.
EPA has expressed its commitment to ground water protection and the value of the
resource through its 1991 policy statement, "Protecting The Nation's Ground Water, EPA's Strategy For The
1990s." Senior officials from all of the Agency's ground water related programs developed this strategy.
Establishing relative priorities is an essential element of this strategy. The strategy states
that "The overall goal of EPA's ground water policy is to prevent adverse effects to human health and the
environment and to protect the environmental integrity of the nation's ground water resources." In
determining appropriate prevention and protection strategies, EPA will also consider the use, value, and
vulnerability of the resource, as well as the social and economic values. EPA's ground water protection
Strategy emphasizes the state role in ground water protection. EPA will soon release final guidance for the
development of comprehensive state ground water protection programs. These comprehensive programs will
be resource based and will provide the framework to coordinate all ground water related programs and
activities under federal, state, and local statutes.
As EPA moves ahead with its state comprehensive programs, making these ground water
valuation judgements will be critical to the states' decisions for protecting this vast resource. This will be
true whether the value is monetized or not. Priorities for action will have to be established and these
priorities will have to be based, at least partially, on the state's perceived value of the ground water.
Today we will begin the dialogue needed to help us develop an approach. There are a
number of questions that need to be considered before we can set values for ground water.
One approach that has been proposed is to value ground water simply as a commodity.
However, if we use price as its value, do we determine a consistent price? A number of factors affect the
price of ground water including location, population served, demand, quantity, and quality. The American
Water Works Association has determined that the average residential water cost in the northeastern United
States is $2.00 per thousand gallons. However, in the Northwest it is about half of that figure, $1.13 per
thousand gallons, due to many of these people receiving subsidized water through the Bureau of
Reclamation.
There are also other questions to be answered. Does valuing ground water via commodity
price provide its total value or should we factor in its other uses? Should ground water be valued the same
whether it is being used for drinking water, irrigation, or industry? We are becoming increasingly aware that
ground water plays a significant role in the hydrologic cycle through ground water/surface water interaction.
GW1-3
-------
Session GW-1
Elder
What is the value of the ground water that recharges surface water and what about the value of maintaining
the ecology of a stream?
Other characteristics of ground water affect its value. Typically, ground water moves slowly.
What we do to ground water will be of concern to future generations, whether we are maintaining its current
quality or cleaning it up. Certain aquifers can provide large volumes of water in short timeframes, such as
sand and gravel surfacial aquifers of the north-central and northeastern United States. Since fluids can move
more easily through them because of the large pore spaces, they are also more easily contaminated if
precautions on chemical use are not taken. It is not easy nor inexpensive to determine the extent of
contamination once it reaches the aquifer and is transported by it. We have learned that such problems may
be with us for years. A more extensive concern is the result of our national pesticide survey, which indicates
pesticides in rural water wells, once thought to be clean. About 10% of the community system water wells
contain one or more pesticides and over 50% contain some level of nitrates. These numbers are not yet
alarming but they do indicate the need to take precautions.
Another issue is, should ground water have the same value for both prevention and
remediation? At EPA, we are charged with preventing the contamination of ground water and with cleaning
up any contamination of the ground water. How much are we willing to pay to clean up the ground water?
The average Superfund cleanup is estimated at over $30 million. What does this suggest about the value of
preventing contamination?
Protection decisions can also be based on the perceived value of the ground water resource.
This perceived value can be based on vulnerability or quality of the ground water or the availability of an
alternative supply. Are pristine aquifers more valuable than slightly contaminated ones? What if it is the
sole source of drinking water for the community? If an alternative supply of drinking water is not available,
ground water may then be priceless.
These are all important questions and many more should be raised. EPA and the other
federal agencies need to formulate sensible responses based on a strategic, long-range view of ground water
resource value and the cost of not responding to contamination. I hope this conference can go a long way
toward developing answers.
GW1-4
-------
Session GW-1
Smith
Valuing Ground Water Resources; A Conceptual Overview
V. Kerry Smith1
Professor
Resource and Environmental Economics Program
North Carolina State University
ABSTRACT
Ground water resources have economic value because people are willing to pay for the
services they provide. This paper outlines a general framework for deriving measures of these values. After
describing how the services of ground water resources can be characterized in an economic model, we can
identify how defining the total value depends on whether an ex ante or ex post perspective is adopted and on
whether use and nonuse values are to be separated. The second section outlines the conceptual logic
associated with the indirect and direct approaches to nonmarket valuation and how they relate to what can
be measured for ground water resources. The last section uses a few examples from recent literature to
illustrate some questions that can arise with attempts to measure the total value for ground water.
1.0 CONCEPTUAL FRAMEWORK FOR ECONOMIC VALUES
The economic framework for defining monetary measures for anything originates in
consumer choice. It is anthropocentric in the sense that people provide the basis for how economic value is
derived. Formal definitions are usually traced to Hicks's7 definition of consumer surplus. However, before
describing the Hicksian framework, it is important to acknowledge that these definitions are derived from an
abstract "thought experiment" on the part of economic analysts. To use them in meaningful terms, we
assume that people seek the highest well-being (or utility) possible within the constraints they face.
Taking that decision framework (i.e., realize the greatest well-being within constraints),
Hicksian definitions are based on the functions that describe the realizations of this process. If we transform
the decision situation by adding more constraints or if we change the context of decisions (e.g., people can
adjust to particular types of uncertainty), then we can have one of two effects. First, expansions in the set of
constraints assumed to condition what people can do will increase the number of factors that could influence
the Hicksian consumer surplus measure. Second, and equally important, if we decide to change the decision
context, this can mean that we are actually changing what is meant by individual well-being. While neither
action alters the logic, both change the definition of Hicksian consumer surplus measures and, as such, affect
what measurement methods can capture.
Consider the simplest statement of Hicksian consumer surplus. People make choices that
are in their best interests. So our valuation thought experiment offers some commodity to people with a
one-tune lump payment, say W, that must be made to obtain the commodity involved. Using the functions
describing how a person's best options within constraints (indirect utility functions) change with the features
of those constraints, we need only substitute the terms into the function, compare the outcome with the
status quo, and (in principle) we can describe choice.
University Distinguished Professor, Resource and Environmental Economics Program, Departments of
Economics and Agricultural and Resource Economics, North Carolina State University. Partial support for
this research was provided by NSF grant number SES-8911372.
GW1-5
-------
Session GW-1
Smith
Of course, in practice the thought experiment is motivated by economists' attempts to go
backward from choices to preferences. So it is hardly surprising that the framework underlying conventional
models readily accommodates a choice prediction. However, suppose we take the logic a step further. This
entails asking a related question. When (at what lump sum) would choice of any kind be unclear? That is,
the individual finds himself (or herself) indifferent. Choice neither increases nor diminishes well-being. At
this point, we define the largest amount a person would pay for whatever is offered. Formally, with indirect
utility functions (i.e., V(P, y, Q), with P = price, y = income, Q = a nonmarketed commodity), we have the
Hicksian willingness to pay, W, for an increase in Q by dQ in equation (1):
y, - W, Q0 + dQ) = Wv yw
-------
Session GW-1
Smith
expected utility constant, and we have nonuse values that may or may not overlap with option values
(Smith19).
The distinction between left and right sides of the figure arises from whether or not
uncertainty is important for people's behavior. If it is and people's responses to the lotteries defined by -K_
are important to their choices of goods and services, then our well-being concept is different (i.e., with the
uncertainty it would correspond to the expected value of utility, while without it, well-being would correspond
to the level of utility).
This increased complexity is important to valuing ground water resources because the
commodity Q must be defined. Various authors have used risk, physical measures, or private consumption
services (e.g., water). Introducing r, q, Q, andw acknowledges that multiple definitions are possible and
there is no reason to assume they will lead to the same values. Indeed, an important aspect of judging the
values recovered by any method will be the consistency between the analyst's treatment of the commodity
and how people interpret it.
With this background, we can turn to the logic underlying the nonmarket valuation methods
used to attempt to measure the components in Figure 1.
2.0 MEASURING NONMARKET VALUES*
2.1 Indirect Approaches
As noted in the previous section, the definitions of Hicksian value concepts begin with
revealed preference analysis. The approach for measuring these values based on choice also relies on the
idea that an individual's selection of a consumption bundle of marketed goods (given their prices) conveys
information. The bundle that a consumer with given income purchases must (because of its selection) be
preferred to all others at that particular set of prices. Because the objectives of indirect approaches for
measuring values extend beyond deriving restrictions to conventional demand functions to recovering
valuation information, more detailed assumptions are required.
To appreciate the features of the indirect approaches, we need do little more than focus on
the marginal rate of substitution (MRS) between the nonmarketed environmental service and some
numeraire as an economic measure of an individual's real value for the last unit he or she consumed of a
commodity. Consider the exchange of goods on markets. Under these conditions, we know that each
commodity's relative price reveals the consumers' real (marginal) values. In developing measures of people's
values for goods and services, the focus shifts to the typical (or representative) individual and the fact that
real values are known once we have the relative prices associated with the consumption bundles selected. If
the amounts purchased are also known, then with sufficient variation in these pairs (i.e., prices and
quantities), we can develop conventional valuation measures. That is, Hicksian measures of consumer
surplus for single price changes (see Hausman6 and Vartia23), and in some cases for multiple price changes
(see LaFrance and Hanemann11), can be recovered from Marshallian demand functions derived from this
information.
The travel cost recreation demand model is the most straightforward of the indirect
methods. Initially proposed in 1947 by Harold HoteUing, the model recognizes that visitors to a recreation
site pay an implicit price-the cost of traveling to it (including the opportunity costs of their time). By
observing an appropriate quantity measure and these costs for individuals at different distances (along with
any entrance fees and related charges), we obtain information comparable to that provided by market
transactions. Of course, this description assumes that the implicit price can be treated as a parameter and
that the quantity measure is straightforward. Once either of these assumptions is modified, the
correspondence to marketed goods becomes less direct.
GW1-7
-------
Session GW-1
Smith
£
E
-
"g cr
* rf
es a
*o a
I I
•S 'c
I
w
1
'63
I
H t,
.. ^
"''"
|
a *• *
ftk. >
II II "
>« >»
GW1-8
"5 -8*
l
B" <*
X IB"
O ^
. .v
»>.
IN
i "I*
^5?
-------
Session GW-1
Smith
Complexities aside, most evaluators conclude that the travel cost methodology has "worked
well." (See Bockstael, McConnell and Strand4; Smith20; or Ward and Loomis24.) What does this really
mean? I believe it is based on four features of the hundreds of estimates accumulated using the travel cost
model. First, the estimates uphold rudimentary predictions of consumer theory (such as having the quantity
negatively related to the own price). Second, when the model is applied to comparable sites, the estimates
reveal a broad consistency in the relative size of price and income elasticities (as well as in the estimated
consumer surplus per unit of site usage). Likewise, estimates across different types of recreation sites reveal
plausible differences in these economic characteristics. For example, we would expect wilderness areas to
behave more like luxury goods than fresh-water boating sites, and this is what the estimates imply. Demand
functions for recreation sites in areas with numerous substitute facilities are more elastic than those for sites
with few comparable alternatives.
Third, meta-analyses of both surplus per unit of use and price elasticities indicate that
modeling judgments affect the estimates as theory would suggest. Moreover, both measures vary with the
type of recreation site being studied. While some observers might argue that the first finding weakens the
method, I do not agree. When a method relies on connecting decisions to underlying preferences using prior
assumptions about what motivates and constrains people's behavior, we should expect these judgments to
matter. What is important is that the impact of each judgment agrees with a priori expectations and that we
can develop methods for testing the assumptions and judgments that are most relevant to recreationists'
circumstances. Finally, controlled experiments suggest that when the assumptions underlying the model are
upheld, the model does perform reasonably well in characterizing underlying consumer preferences.b
A second class of indirect valuation methods uses averting behavior or household production
function (HPF) models to infer an individual's value for some aspect of environmental quality when private
actions can influence how it is experienced. Like Retelling's insight in the travel cost model, the HPF
framework uses people's actions to isolate features of their values. In this case, the choices observed in the
HPF framework involve reallocating expenditures on marketed commodities or time to adjust to a change in
the amount of some nonmarketed resource. The HPF framework alone does not add information. Rather,
it offers a rationale for imposing restrictions on preferences (or what is referred to as the household
technology) so observable decisions can provide the necessary valuation information.
A number of different applications unking a pollution measure to some observable response
are classified as versions of the HPF method. Physical damage functions may well be the most common. A
second empirical model often suggested as based on the HPF framework involves models for reported
mitigating behavior, such as purchasing a filter to remove water pollutants (see Smith and Desvousges22 or
more recently Abdalla et al. ). Neither of these provides sufficient information to recover valuation
information.
Of course, with assumptions, HPF applications can be restricted to exhibit certain
properties. For example, perfect substitution underlies using expenditures for a commodity that is assumed to
be a perfect substitute for the services of an environmental resource to value that resource. Another
possibility involves maintaining the equivalent of weak complementarity by assuming that one factor input is
an essential input in the household production technology (Bockstael and McConnell3). Other combinations
of features of the HPF also can be interpreted as functional restrictions imposed on consumer preferences.
1 f*. 10
(See Neill and Larson as examples.) A unifying principle connects each of these strategies: taken
together with a general specification for preferences, they restrict the factors motivating individual choice so
the desired MRS can be linked to an observable set of relative prices (whether actual prices or implicit
costs). [Emphasis added by author.]
The marginal rate of substitution is also the preference information recovered from the last
type of indirect method to be reviewed-hedonic price functions. By specifying characteristics that distinguish
heterogeneous commodities and by recognizing that equilibrium means an absence of incentives for
arbitrage, the hedonic model describes how a set of prices for each of the commodities will define the
GW1-9
-------
Session GW-1
Smith
equilibrium. With a large enough number of different commodities, this equilibrium is characterized by a
price function that describes how prices change with the characteristics of the commodities involved. This
insight can be used in nonmarket valuation because some of these characteristics are site specific. Market
participants must be aware of this site connection and share a common basis for recognizing it if these
characteristics are to influence the prices.
22 Direct Approaches
Direct valuation approaches, usually designated as the contingent valuation methods (CVM),
replace observed behavior with survey responses about values, stated choices, ranks, matches, or other types
of behavior. There are three important differences between most indirect methods and CVM.
The most significant difference arises from the ability to control the commodity presented to
people as an integral part of the survey because what is asked is hypothetical. This characteristic is both a
strength and a weakness. When the features of the commodity, conditions of access, payment format, and all
relevant details of what is presented to the individual are controlled, there is no need to rely on analyst
judgments in adapting a scenario or constructing a variable (or both) to represent the commodity or access
conditions.
A second distinction follows from the requirement to collect primary data. It can ensure
better targeting of the populations likely to be affected by the injuries. While it would be possible to
undertake new indirect surveys to provide comparable information as the CVM application, this is rarely
done. As a consequence, data from the best CVM studies generally have richer detail on socioeconomic
characteristics and attitudes and multiple ways of cross-checking responses.
Finally, given the assumption that responses to hypothetical questions are consistent with
real behavior, then these responses usually have a more direct link to Hicksian measures of consumer surplus
than those developed from the indirect methods. Nonetheless, the hypothetical nature of the responses has
remained controversial with early issues concerned about strategic reactions and more recent work focused
on the hypothetical nature of the question themselves. Because it appears relatively straightforward to
conduct CVM studies, a larger number of studies could be judged as poorly constructed.
Because mainstream economists remain skeptical of the insights derived from people's
responses to hypothetical questions, initial objections to CVM crystallized around two general questions: is
CVM reliable, and is it accurate? The next stage in CVM research sought to address these questions in a
variety of ways. The current stage in CVM research is a derivative of both of these efforts. It recognizes the
need to integrate the psychological and economic dimensions of framing CVM questions with conventional
practices in survey design, implementation, and analysis. (See Mitchell and Carson15 for a comprehensive
treatment.)
What conclusions can we draw from comparing CVM with other alternatives? I believe the
record indicates that some forms of CVM can provide theoretically consistent and plausible measures of
individuals' values for some types of environmental resources. The types of commodities need not be limited
to the narrow set defined by Cummings, Brookshire, and Schultze's5 reference operating conditions.
However, we are far from identifying how the characteristics of the commodities to be valued, the attributes
of the people whose values are to be measured, and the features of the survey influence the reliability of
CVM estimates.
3.0 APPLYING THE METHODS FOR GROUND WATER VALUATION
Research with specific estimates for ground water valuation is quite limited. While
examples can be identified using both the indirect and direct approaches in the early literature, the individual
studies do not appear to offer convincing estimates. The evidence for values of ground water resources
based on hedonic studies is exceptionally limited. For example, Michaels and Smith14 and Kolhase10
GW1-10
-------
Session GW-1
Smith
investigated hedonic property value models with distance to hazardous waste sites as proxy variables for
disamenity effects. While some of the communities in the Boston sample used by Michaels and Smith had
water supplies from ground water, in this case distance would not be serving as a proxy for this exposure
channel.
There are at least two reasons for this conclusion. The towns involved had wells that were
at different distances from the hazardous waste sites, and they generally were not related to the proximities
measured for individual houses. Equally important, water supplies come from multiple sources, so it is
unreasonable to assume people would know the specific details of its provision.
Comparable issues can be raised with the Kolhase application. Moreover, even if ground
water were the only exposure channel, these models would reflect an ex ante use value based on the risk of
contamination perceived from hazardous waste sites at varying distances from homes with ground water-
based water supplies.
Averting cost studies are equally limited, with the literature providing some evidence of a
link but little of direct use for valuation. (Smith and Desvousges22 and Abdalla et al. provide examples.)
The more recent study by Abdalla developed estimates of aggregate averting expenditures but were not
successful in Unking these costs responses to the risk of contamination of ground water because less than half
(43%) of the households in an area with TCE-contaminated ground water were aware of it after a mandated
notification.
Recent contingent valuation studies (see Powell18 and Poe and Bishop17) suggest that it
may be possible to use protection of ground water from contamination (largely to avoid household
consumption related impacts) as the basis for valuing some of the services people associate with ground
water resources. However, neither study is specific about what the ground water protection programs offer.
Their commodities are composites of reduced risk of contaminated ground water, along with quality
characteristics that are left to each individual to define. Thus, a lottery that consists of n, R, and r is
presented to each person and we are not sure of the relative importance of each element. This is not
important for each study because the commodity doesn't vary across the respondents in each survey. It is
important to our ability to transfer their findings to new applications.
Overall, then, it appears that limited evidence is available on the performance of using
indirect and direct methods to value ground water resources. What is available seems to indicate that people
can understand questions about protecting ground water and can state hypothetical values for doing so.
However, it is not clear how these values would relate to their total valuation of ground water resources.
4.0 NOTES
a. This discussion draws upon a longer review of nonmarket valuation literature (see Smith21).
b. Kling's research ' provides an ideal example of such evaluations. More recently,
Adamowicz et al.2 examined the sensitivity of consumer surplus estimates to mis-
specifications of the travel cost demand function.
5.0 REFERENCES
1. Abdalla, C.W., BA. Roach, and DJ. Epp. "Ground Water Contamination." Land
Economics. Vol. 68, pp. 163-169. May 1992.
2. Adamowicz, V.L., G.L. Fletcher, T. Graham-Tomasi. "Functional Form and the Statistical
Properties of Welfare Measures," American Journal of Agricultural Economics. Vol. 7, pp.
414-421. May 1989.
GW1-11
-------
Session GW-1
Smith
3. Bockstael, N.E., and K.E. McConnell. "Welfare Measurement in the Household Production
Framework," American Economic Review. Vol. 73, pp. 806-814. 1983.
4. Bockstael, N.E., K.E. McConnell, and I. Strand. "Recreation," Measuring the Demand for
Environmental Quality. 1991.
5. Cummings, R.G., D.S. Brookshire, and W.D. Schultze. Valuing Public Goods: A State of
the Arts Assessment of the Contingent Valuation Method. Totowa, NJ. 1986.
6. Hausman, J A. "Exact Consumer's Surplus and Deadweight Loss," American Economic
Review. Vol. 71, pp. 662-76. September 1981.
7. Hicks, J.R. "Consumers' Surplus and Index Numbers," Review of Economic Studies.
Vol. 9 pp. 216-137. 1941-42.
8. Kling, C.L. "Comparing Welfare Estimates of Environmental Quality Changes from
Recreation Demand Models," Journal of Environmental Economics and Management. Vol.
15, pp. 331-40. September 1988.
9. Kling, C.L. and M. Weinberg. "Evaluating Estimates of Environmental Benefits Based on
Multiple Site Demand Models: A Simulation Approach," Advances in Applied Micro-
Economics. JAI Press, Greenwich, Ct. Vol. V. 1990.
10. Kolhase, J.E. "The Impact of Toxic Waste Sites on Housing Values," Forthcoming, Journal
of Urban Economics. 1992.
11. LaFrance, J.L. and W.M. Hanemann. "The Dual Structure of Incomplete Demand
Systems," American Journal of Agricultural Economics. Vol. 71, pp. 262-274. May 1989.
12. Larson, D.M. "On Measuring Existence Value," Forthcoming, Land Economics. 1992.
13. McConnell, K.E. 1983. "Existence and Bequest Value," Managing Air Quality and Scenic
Resources at National Parks and Wilderness Areas. Westview Press, Boulder, Co. 1983.
14. Michaels, R.G, and V.K. Smith. "Market Segmentation and Valuing Amenities With
Hedonic Models: The Case of Hazardous Waste Sites," Journal of Urban Economics.
Vol. 28, pp. 223-242. 1990.
15. Mitchell, R.C. and R.T. Carson. Using Surveys to Value Public Goods-The Contingent
Valuation Method. Resources for the Future, Inc., Washington, D.C. 1989.
16. Neill, J.R. "Another Theorem on Using Market Demands to Determine Willingness to Pay
for Non-Traded Goods," Journal of Environmental Economics and Management. Vol. 15,
pp. 224-232. June 1988.
17. Poe, G.L., and R.C. Bishop. "Measuring the Benefits of Ground Water Protection From
Agricultural Contamination: Results From a Two-Stage Contingent Valuation Study," Staff
Paper No. 341, Department of Agricultural Economics, University of Wisconsin. 1992.
18. Powell, J.R. The Value of Ground Water Protection: Measurement of Willingness-to-Pav.
Information and Its Utilization by Local Government Decisionmakers. Unpublished Ph.D.
thesis, Department of Agricultural Economics, Cornell University. 1991.
GW1-12
-------
Session GW-1
Smith
19. Smith, V.K. "Nonuse Values in Benefit Cost Analysis," Southern Economic Journal. Vol. 54,
pp. 19-26. July 1987.
20. Smith, V.K. "Taking Stock of Progress with Travel Cost Recreation Demand Methods:
Theory and Implementation," Marine Resource Economics. Vol. 6, pp. 279-310. 1989.
21. Smith, V.K. "Nonmarket Valuation: An Interpretive Appraisal," Forthcoming, Land
Economics. February 1993.
22. Smith, V.K. and W.H. Desvousges. "Averting Behavior: Does it Exist?" Economic Letters.
Vol. 20, pp. 291-%. 1986.
23. V^rtia, Y.O. "Efficient Methods of Measuring Welfare Changes and Compensated Income
in Terms of Orderly Demand Functions," Econometrica. Vol. 51, pp. 79-98. January 1983.
24. Ward, FA., and J.B. Loomis. "The Travel Cost Demand Model as an Environmental Policy
Assessment Tool: A Review of Literature," Western Journal of Agricultural Economics.
Vol. 11, No. 2, pp. 164-178. 1986.
GW1-13
-------
Session GW-2
Introduction
Session GW-2: Methods and Case Studies - Recent Work
SESSION SUMMARY
MODERATOR: V. Kerry Smith
PRESENTERS:
William Schulze— Do Existence Values Exist for Ground Water? Methods and Case
Studies
John Bergstrom-.Bene/ifs of Protecting Ground Water from Agricultural Chemical
Contamination
Charles W. Abd&llar-Avoidance Costs and Ground Water Values: Results of Two
Empirical Applications
Erik Lichtenberg-lfrmg Risk Assessment in Analyzing Ground Water Protection
Policies
Gregory Michaels- When the Home is No Longer A Castle: Inferring the Economic
Value of Ground Water Contamination from Residential Property Sales
GW2-1
-------
Session GW-2
Schulze
Do Existence Values Exist for Ground Water?
Methods and Case Studies
(transcribed from audio tape)
William Schulze
Department of Economics
University of Colorado
This study was sponsored by the Office of Policy Analysis and aided by the Office of Solid
Waste. It focused on non-use values for cleaning up contaminated ground water. It is rather a difficult
question indeed and I proposed this question. Existence values are the values of knowing that an
environmental resource is preserved or protected even if no one was ever to use it. The source of non-use
values is bequest value, which is the desire to leave the natural resources for future generations.
The people who did the work in addition to myself are: Jeff Lazo, research assistant in the
Economics Department; Gary McClellon, psychologist; and Jim Doyle, research assistant in the Psychology
Department associated with the study.
The study was the result of a series of studies that have been funded by EPA at the
University of Colorado in which there has been a collaborative effort between psychology and economics, in
which we have attempted to determine what constitutes an acceptable contingent valuation or survey study
for determining environmental values. This was necessary to allow measurement of the use versus the non-
use values. So in this case, that would be the difference between consumptive use of water, which is fairly
easy to get the value for, and non-use values, such as existence and bequest values, which are quite difficult
to determine.
We first did two studies of familiar commodities, commodities that are relatively easy to
value using the survey approach. The first is a study of air pollution in the Denver area. The second is a
study of eastern visibility. The second major study that we did was to look at what economists are now
beginning to call an "exotic commodity", namely ground water clean up.
People are obviously not as familiar with ground water cleanup. I use the term "commodity"
in the sense that economists use it. Ground water clean-up is just like buying a hamburger. You buy ground
water clean-up. The difference is that you are familiar with a hamburger, you are not familiar with ground
water clean-up. So the question is, what happens if we ask people about the value of unfamiliar
commodities?
In this research, we used a new tool referred to as cognitive survey design. I will describe
that to you. It is a tool drawn from psychology. People who design surveys have been shocked in recent
years to find out that when people answer questions asked in surveys, they very often are getting answers
different from the ones that they thought they were getting.
And the way they found this out was literally to stick a microphone in the face of people
and force them to speak continuously while they are answering the survey. So you get kind of a window into
their thought processes. And that can be very revealing. In fact, we will look at some verbal protocols today
to see, if in our survey about ground water values, we were getting the kinds of answers we initially thought
we would get. And, in fact, I will show you that in some cases, we did not.
The second is this concept of value construction that Kerry Smith began to mention. And
that is certainly for something like ground water value as opposed to the value of a hamburger. People do
not have that value just sitting in their head. You cannot go up and ask them a question like "How tall are
you?" When you know how tall you are, you can just answer it easily.
GW2-2
-------
Session GW-2
Schulze
If I ask you "How much would it be worth to you to clean up contaminated ground water in
your community if you have that problem?" That is not something that you can just roll off your lips and
give me an answer to it. You have to think about it. In effect, you have to construct a value for me. And
the value you construct may well depend on how much you know about the problem.
I can summarize the results of this series of studies as follows. For familiar commodities,
there was little impact of the information in the context provided in the survey. In the case of these air
quality studies, what turned out to be true was that people had a good understanding of air quality issues.
And when we gave them more or less information, it had variable impact on their values.
In terms of the ground water study that I really am going to talk about in just a moment, it
turned out that information in the context provided in the survey, had a very dramatic impact on the values
that we received. In other words, people did not know much about ground water and educating them about
ground water was quite important in determining their responses.
What this suggests in terms of implications, is that surveys would attempt to estimate non-
use values plus really provide perfect information. How do you determine what information is needed?
Everything? What could you leave out? Anything you leave out might bias the survey. So that is a rather
daunting task in the process of survey design.
And similarly, that these surveys must provide a complete transactions context. If you are
going to get people to respond in a monetary way, they have to actually believe that this is a monetary
transaction. For example, in the ground water clean-up study, we proposed that people would pay an
additional fee on their monthly water bill. That is a fairly believable way of paying for ground water clean-
up for response.
So the question is, how can we handle the seemingly impossible burden on survey design,
which is to provide perfect information, to make that information inherently unbiased. So the procedure that
we developed or, at least, an idealized procedure which we would propose, is first to find out what
information is relevant. One would really like to use a panel of experts. They could develop the kind of
information that people ought to have if they are going to give you values for ground water clean-up. Now,
OSW performed that service for us in the case of this study.
The second step in this process is to develop a perfect information complete kinds of
contact survey instrument. Now this survey instrument does not look anything like a survey that you could
actually use in the field. In the case of our preliminary design, it was 22 pages long, single spaced, with lots
of diagrams and risk ladders.
We paid people, basically, to go through this exercise and to become experts in ground
water before giving us the values. So we take this rather lugubrious approach. We then test this method in
looking for various problems by using this approach where we ask people to fill out this 22 page instrument.
We have a microphone in front of them. We ask them to go through the whole thing and
we hear their thoughts because the process encourages them to verbalize. There is no cattle prod involved.
There is just a research assistant encouraging them and they never stop talking. We have to pay people a lot
to get them to do this.
And we look for problems like imbedding. Are they really giving us values for more than
what we are asking for? Very often people have their own thoughts and will value something differently
from what you want them to value. Do they reject the scenario? Do they just not believe the material that
you are giving them? Do they say, "That's a stupid idea," and bid zero, not because they do not value the
environmental clean up, but do they bid zero because they just do not like the way you were doing it. That
is called scenario rejection.
GW2-3
-------
Session GW-2
Schulze
Are they understanding the information that you are providing in this survey instrument?
At that point, after the verbal protocols, you redesign the survey instrument. Then you go to groups of about
50 or 100. Give them this still lengthy survey, but you debrief them and ask them, "What information did
you use in constructing your value?" Based on those responses, you throw out things that you might have
thought were very important, but people just do not care about.
And after discarding the unused information and material, you see if you have a survey that
is short enough to go out in the field with. In the case of this ground water study, and actually somewhat
surprisingly, we found we could get this thing down to a 12 page mailable survey by focusing just on the
information that people told us they wanted. And, in fact, before doing that, we tested with a sample of 100
people to make sure that the shortened survey gave essentially the same values as the long, full information,
full context survey.
To give you an idea of what was in that full information, full context survey, we went
through a nine step process in those 22 pages plus boxes, arrows and diagrams where we first gave them
some ground water education. We showed them a risk ladder. We asked them to evaluate their response to
a water shortage. We might not clean up the ground water, you might just use less water. The idea was to
get them to value water itself in terms of its use.
We had them value an option of buying water from another city as a temporary transfer, a
choice of in-home water purification. We had them value a fund for future use. You could leave money and
not clean up your own water, but simply put money in the bank for 50 years. And if you put $1.00 in the
bank for 50 years, it is worth $100 at a reasonable interest rate. You could just leave money for future
people to clean up the problem.
We have them value public water supply treatment and then evaluate complete ground
water treatment, which was a pump-and-treat option. And then we asked them questions about, "Are these
values really just for ground water clean-up?"
What kind of mental models did people have of ground water? We all know that ground
water moves maybe 100 feet a year under ground. But here is what people think and it is quite important in
this construction of values. These are off the tapes: "Probably not very fast." "It probably depends on where
the water comes from, two feet per second." "In two hours-it might go 10 miles in two hours."
Now think. If you have contamination under ground, you are going to believe that that
contamination is going to be all over your state in a few days. And there are people that actually believe
this. Is that going to raise or lower your value for ground water clean-up? Obviously, it is going to raise it.
"Thirty miles per hour tops. It shoots out of there pretty quickly."
"I would say pretty quick—like in miles per hour. It has to be quicker than people would
guess. It is not nearly as quick as a river, but I know that it flows out of the fields. It seems it could go
through a mile in the matter of an hour. If the water is moving that fast, I would have to guess that it was
something that is fairly shallow, like a city water supply, it could go, at most, maybe 10 or 15 miles and so
on."
There is one down there that says, "Not surprised. I thought it moved slower. I had a
geology class recently, so I was aware of how ground water works and functions." So we found one person
that knew something about ground water. Now the interesting thing was, people who did not know about
ground water, kind of realized they did not. And so they accepted the information in the survey. They
actually did not reject the information we provided to them. But, obviously, this is one thing that is going to
affect values.
The idea of a trust fund, this notion of, "Well, gee. Let's just leave money to solve this
problem in the future", was completely rejected by most or at least many of our respondents. For
GW2-4
-------
Session GW-2
Schulze
example-"! don't care. I am rejecting your scenario. No, just in the sense that-I don't know. It would be
there. They might spend it on something else. When are they going to dip into it to use it? I don't know."
The local government or people may want to dip into this fund and so on. It is like freezing
your body to see if there is something in the future to handle it. I am not a big believer in that. I would like
to believe it, but when they start talking about the S & L scandal, I don't know.
The worth of $1.00 in a bank in 50 years is probably 10 cents. I do not think it would be
there the way my bank has service charges.
That is just not an option. And you thought that survey design was not going to be fun. So,
this option is gone. We could not use it as a way of getting at values. The top chart is a frequency
distribution of the values for people who did not get all this information just to see if information made a
difference. The thing you cannot see very well is that is a logarithmic horizontal axis.
The second chart of frequency distribution shows what happens after people get all of this
information. There is a collapse in the variance. But, more importantly, because that is a logarithmic,
horizontal axis in terms of dollar values that people gave us, there is a tremendous collapse hi the mean from
about $20 a month down to $12 a month.
We went to a national survey of 5,000 people. We got a 63 percent response rate. We used
three methods to try to get at non-use values. The first was simply to ask, "Of your total value for cleaning
up the ground water, which would also provide water for use, what percent was for non-use?" That is,
existence and bequest values. And then take that percent of their total value.
The second approach was to use scenario differences. We asked them, "What is the value
for complete ground water clean-up? What is the value you would pay for just surface water treatment?"
The difference, in essence, in those two values is a value for just cleaning up under ground for existence and
bequest values, since complete clean-up guarantees that water is available for the future as well as just
knowing that it is cleaned up under ground. And people cared about those things.
Finally, we varied the magnitude of the water shortage in the scenarios from 10 to 40 to 70
percent. That allows us to extrapolate back to zero percent and see if there is any interest out there. Is
there some value even if the shortage is zero percent?
In these three approaches, we got for the percent splits, $3.49 a month. For the
extrapolation value, we got $3.54 a month, so those are statistically similar. The one dissimilar one is
scenario differences, which is the difference between the willingness to pay for complete clean-up and the
willingness to pay for treatment at the surface, as you pump it up.
And so, if you built that treatment plant that might partly cover the bequest values, so we
would expect this scenario difference maybe to be an underestimate. I may just be rationalizing. But those
numbers are at least close, which demonstrates that once people have the information, at least you can get
consistent values out of them using different ways of attempting to measure a value.
If they do not have the information, what happens is it is easy to show the inconsistent
responses. But you have to provide the information to get logically consistent, and I think, meaningful
values.
GW2-5
-------
Session GW-2
Bergstrom
Benefits of Protecting Ground Water from Agricultural
Chemical Contamination
John C. Bergstrom3
Associate Professor
Department of Agricultural Economics
University of Georgia
ABSTRACT
The results and implications of a study designed to measure the benefits to the general
public of protecting ground water from uncertain future agricultural chemical contamination are discussed in
this paper. The theoretically appropriate measure of ground water protection benefits under demand and
supply uncertainty is option price. Option price associated with ensuring the protection of ground water
from agricultural chemical contamination was measured for a sample of registered voters in Dougherty
County, Georgia using the contingent valuation method. Significant determinants of option price included
annual income, age, subjective concern over the effects of pollution on one's health, and subjective supply
uncertainty. Further testing of the contingent valuation method as a means for valuing ground water quality
is recommended.
INTRODUCTION
A number of recent studies suggest that ground water supplies in many regions of the
nation are threatened with contamination by agricultural chemicals.1'2 In response to this threat, government
agencies are considering various strategies for protecting ground water supplies from agricultural chemical
contamination (e.g., U.S. EPA, 1987). Evaluation of these strategies may be facilitated by estimating the
economic value to the general public of ground water protection.
The purpose of this paper is to summarize the results and implications of a recent study
designed to estimate the economic benefits of ground water protection to citizens of Dougherty County,
Georgia. The general valuation problem is summarized first. Next, the methodology for estimating ground
water protection benefits is discussed. The results of the valuation study are then presented. Implications of
the results are discussed in the final section.
1.0 VALUATION PROBLEM
Dougherty County is located in the southwestern region of the State of Georgia in the
southern Atlantic Coastal Plain. This region is underlain by a deep succession of sand, clay, and carbonate
rocks which form a large aquifer system. Ground Water from this aquifer system is the source of almost all
public and private drinking water supplies in Dougherty County.
Because of the availability of abundant ground water, good sandy soil, and a mild climate,
agriculture is one of the largest industries in southwest Georgia. Agricultural production in the region
involves heavy use of chemical fertilizers and pesticides. The combination of this heavy chemical use with
the local geologic structure (e.g., relatively porous soil) creates a high potential for seepage of pesticides and
aJohn C. Bergstrom is an Associate Professor in the Department of Agricultural and Applied Economics,
The University of Georgia, Athens. The contributions of Jeff Dorfman and Henglun Sun with respect to
data collection and analysis are gratefully acknowledged.
GW2-6
-------
Session GW-2
Bergstrom
nitrates from fertilizers into ground water supplies in Southwest Georgia, particularly the Dougherty County
area5-6-1).
According to the most recent ground water quality tests available, ground water quality in
Dougherty County is currently "safe."7-1-2 In this case study, the general valuation problem was therefore to
estimate die benefits of protecting the current "safe" level of ground water quality from potential future
contamination. The valuation problem is illustrated by Figure 1. In Figure 1, chemical uses refers to human
activities such as mixing chemicals at wholesale and retail farm stores, mixing and applying chemicals on
farms, and disposing of used chemical containers. Physical pathways to contamination refer to different
routes by which agricultural chemicals may enter ground water supplies (e.g., soil, wellhead, sinkhole).
Current Groumrwater
Quality (Q°)
Future Groundwater
Quality (Q1)
Program
Figure 1. Illustration of Valuation Problem
Chemical uses combined with physical pathways create potential ground water
contamination situations. For example, improperly mixing highly concentrated chemicals near an unprotected
wellhead may result in contamination. Contamination may also occur in situations where no negligence is
involved. For example, even if fanners are properly mixing and applying chemicals, contamination may occur
if the soil is relatively porous and the underlying aquifer is relatively close to the surface.
Suppose a program was proposed to protect the current "safe" level of ground water quality
in Dougherty County. As illustrated in Figure 1, an assessment of the program features, future chemical
uses, and potential physical contamination pathways provides information on the "with" and "without"
program probabilities of future contamination. We can then define the "with program" level of expected
ground water quality as:
GW2-7
-------
Session GW-2
Bergstrom
(1) E[QW] = Q°*ZW + Q^l-Z™)
where Q° represents the current uncontaminated ground water quality, Zw represents the probability of no
future contamination with the program, Q1 represents contaminated ground water quality, and (1-Z^
represents the probability of future contamination with the program. Similarly, we can define the "without
program" level of expected ground water quality as:
(2) E[QWO] = Q°*ZWO + Q1*(1-ZWO).
where Zwo represent the probability of no future contamination without the program and (1-ZWO) represents
the probability of future contamination without the program. The expected change in ground water quality
"with" and "without" the protection program is therefore defined as:
(3) E|AQ] = E[QW]-E[QWO] = [Q°*ZW + Q^l-Z")] - [Q°*ZWO + Q1:*(1-ZW0)].
We can now define a bid function for the ground water protection program generally as:
(4) WTP = f(E|AQ],S,X)
where, WTP is a Hicksian equivalent measure of willingness-to-pay for the ground water protection program,
S is a vector of probabilities of different "states of the world" which are independent of the ground water
protection program, and X is a vector of other potential determinants of WTP.
Supply uncertainty influences WTP through the term E[AQ]. WTP may also be subject to
demand uncertainty through the term S in (4). For example, a resident may move out of Dougherty County
sometime in the future for reasons unrelated to ground water quality (e.g., job transfer). WTP is therefore a
measure of option price, which is the theoretically appropriate measure of economic value under supply
and(or) demand uncertainty.8'9
2.0 VALUATION METHODOLOGY
Data on Dougherty County citizens' willingness-to-pay (WTP) for a ground water protection
program were collected using a contingent valuation method (CVM) mail survey. The referendum format
was selected for this application of the CVM.10'11'12 The referendum format involved asking survey
respondents to indicate if they would vote "yes" or "no" for a proposed ground water protection program. A
respondent's "yes" or "no" response was considered a "contingent vote". A "contingent vote" is defined as a
vote which is contingent upon the conditions and outcomes of the referendum constructed in the survey
questionnaire actually occurring.
3.0 QUESTIONNAIRE RATIONALE AND DESIGN
In the questionnaire, survey respondents were asked to assume that the ground water
protection program would ensure the maintenance of "safe" ground water quality into the future. Thus, in
the "with program" situation, the probability of no future contamination (Z*) was 100%.b Survey
respondents were asked to provide a subjective estimate of the probability of future contamination "without"
the protection program (1-ZWO).
bScientific information was not available for estimating the actual effectiveness of a groundwater protection
program. An effectiveness level of 100% was therefore selected to provide a frame of reference which was
relatively simple for survey participants to understand.
GW2-8
-------
Session GW-2
Bergstrom
Let ground water quality be defined by an indicator variable Q such that Q = 1 if ground
water is uncontaminated (within EPA standards) and Q = 0 if ground water is contaminated (violates EPA
standards). Thus, in Equation (3), Q° = 1 and Q1 = 0 which implies that (3) reduces to:
(5) E[AQ] = Z^Z™.
For the Dougherty County case study, the bid function specified in (4) therefore reduces to:
(6) WTP = f(Zpr-Zwo,S,X).
Because respondents were asked to assume that the program would protect ground water with certainty (e.g.,
Zw = 1), (6) further reduces to:
(7) WTP = f(l-Zwo,S,X)
where (1-Z™0) represents a respondent's subjective assessment of the probability of future contamination
without the program.
Thus, the "commodity" or service provided by the protection program was a reduction in the
risk of future ground water contamination equal to (1-Z*0). Respondents were informed hi the
questionnaire that the costs of the protection program would be paid by Dougherty County citizens in the
form of increased grocery prices, tax payments, or other changes which would reduce their annual income by
a certain amount per year. The following valuation question was then asked:
Would you vote to support the program for preventing ground water pollution from
agricultural pesticides and fertilizers, if the program reduces the amount of money you have
to spend on other goods and services by $ per year?
The amount of the income reduction was varied randomly across the sample. The income reduction
amounts used were $5, $20, $70, $100, $150, $250, $350, $500, $1,000, $1500, and $2000.
4.0 SURVEY IMPLEMENTATION
Implementation of the mail survey followed procedures suggested by Dillman's13 total
design method. A pretest of the questionnaire was first conducted on a small random sample of Dougherty
County registered voters. The final questionnaire with a cover letter was then mailed out to a random
sample of 1,440 registered voters in the county. A postcard reminder was sent to all persons in the sample
one week later. Three weeks after the initial mailing, a replacement questionnaire with a second cover letter
was sent to all nonrespondents. The time period of the survey was October-November 1989.
5.0 ESTIMATION OF WTP
The CVM survey generated a series of contingent votes on the ground water protection
program, given the cost of the program in the form of an annual income reduction. There are several
approaches for estimating WTP from discrete voting data. One approach involves selecting a specific utility-
theoretic specification for the indirect utility function, and then deriving a bid function from that
specification.11 Another approach is to select a more ad hoc specification for the bid function, and then
directly estimate the bid function using econometric procedures suggested by Cameron.10
An advantage of the utility-theoretic approach is that a bid function can be derived and
estimated which strictly conforms to consumer demand theory. A disadvantage of this approach is that the
functional form of the estimated bid function may be relatively inflexible. An advantage of the ad hoc
approach is that it allows the estimated bid function to take on more flexible functional forms. A
disadvantage of the ad hoc approach is that the estimated bid function cannot be directly linked to a specific
GW2-9
-------
Session GW-2
Bergstrom
utility function. The estimated bid function, however, can be interpreted as an approximation of a bid
function which could be derived from some specific utility function.
One of the objectives of this case study was to examine the effects of certain independent
variables on WTP (option price) for ground water quality protection. Because of its flexibility, the ad hoc
approach is generally more amenable than the utility-theoretic approach for this type of multivariate analysis.
The ad hoc approach was therefore selected for estimating WTP. The bid function was specified in a semi-
logarithmic form as:
(8) WTP =a +£ ^ogCM) +0 2log(OWN) +0 3log(AGE) +y3 4log(SUPP) +/3 5log(DEMN), where
M = annual income, OWN = index measuring a respondent's concern about the effects of environmental
pollution on his or her health, AGE = respondent's age, SUPP = respondent's subjective probability of
future contamination within 5 years without the protection program (measure of supply uncertainty) and
DEMN = respondent's subjective probability of demanding ground water in the future (measure of demand
uncertainty).
Demand for environmental quality is expected to increase with income (M). It was
therefore expected that y3 j in (8) would have a positive sign. The more concerned a person is about his or
her own health, the more likely is he or she to be willing-to-pay for ground water quality protection. A
positive sign on/32 was therefore expected. Conceptually, the effect of a person's age on preferences for
environmental quality is rather ambiguous. The expectation of the sign on ft 3 in (8) was therefore positive or
negative. An increase in the probability of future ground water contamination is expected to increase the
demand for ground water quality protection. An increase in the probability of demanding ground water in
the future is expected to increase the demand for ground water quality protection. Thus, /34 and f)5 were
both expected to have positive signs.
6.0 VALUATION RESULTS
6.1 Response Rate and Protest Bidders
Of 1,440 surveys sent out, 156 were returned as undeliverable, leaving an adjusted sample
frame size of 1,284. Of this amount, 660 questionnaires were returned for a response rate of 51.4%. This
response rate is quite comparable with those of other nonmarket valuation studies which have used a mail
survey with two follow ups. '
Cummings, Brookshire, and Schultze15 state if a person bids zero as a "protest" to being
asked to pay for an environmental good, the bid is not an indicator of his true valuation. Protest bids are
therefore typically screened out of the sample.16 In this case study, respondents who indicated they had an
inherent "free" right to clean ground water or they refused to place a monetary value on clean ground water
were considered "protest bidders." The total number of protest bidders was 57 which represented 8.6% of
the usable sample.
62 WTP Estimation Results
Upon eliminating the protest bidders, there were 603 valid observations (91.4% of
responses) for the bid function estimation. For these observations, the sample population had an average
age of 46.8. Average years in residence was 23.1. Average education was 13.8 years. Average household
size was 3 with 1 child. Average subjective pollution probability was 54% and average subjective demand
probability was 68%. The average annual household income was $42,517 with a range between $5,000 and
$500,000 (Table 1).
The WTP function in (8) was estimated following the censored logistic regression
procedures suggested by Cameron.10 The coefficients on annual income (M), own health concern level
GW2-10
-------
Session GW-2
Bergstrom
(OWN), and subjective contamination probability (CONT) had expected positive signs which were statistically
significant. The coefficient on subjective demand probability (DEMN) had a positive sign as expected, but
was not statistically significant. The age variable (AGE) had a statistically significant coefficient with a
negative sign (Table 2).
Using Cameron's approach, mean WTP for the ground water protection program was
calculated at $641 annually per household. This mean value was derived using the average values for the
independent variables shown in Table 1. For example, the $641 mean value was derived using the mean
subjective contamination probability (without the control program) of 54%. The 95% confidence interval of
the option price is between $890 and $493. Using data from the 1990 Census of Population and Housing,17
mean household income in Dougherty County was estimated at $34,550 annually and the mean age of the
"head of the household" was estimated at 43 years. Thus, a sample of Dougherty County citizens used to
estimate WTP was skewed slightly towards higher income, older people. The sample also contained only
registered voters. Thus, it is not generally valid to extend the $641 per household estimate of mean WTP to
the Dougherty County population in general.
In order to gain insight on how mean WTP might change with changes in certain variables,
a sensitivity analysis was conducted by changing one independent variable by one standard deviation (or to
the extreme value) and holding all other variables at then- mean values. The results (Table 3) suggest that if
mean annual income increased to $78,000 mean WTP would increase to about $1,450 annually. If mean
annual income decreased to $7,000, mean WTP would decrease to about $165 annually. If the mean age of
the head of the household is 31, mean WTP would be about $870. If the mean age of the head of household
was 62, mean WTP would decrease to $469. If the typical respondent was "very concerned" about the effects
of pollution on his or her own health, mean WTP would be around $905. If the typical respondent expressed
"no concern" about the effects of pollution on his or her own health, mean WTP would decrease to about
$71.
A subjective probability of future contamination (without the program) of 100% would
result in a mean WTP of about $942. If the subjective probability of future contamination (without the
program) was 0%, there would still be a positive mean WTP of about $120. If the typical respondent has a
subjective demand probability of 100%, mean WTP is around $682. This compares to a mean WTP of $451
assuming the typical respondent has a subjective demand probability of 0% (perhaps indicating the presence
of non-use values). As the above results indicate, mean WTP appears to be quite sensitive to changes in
supply uncertainty, demand uncertainty, personal health attitudes, age, and income.
Several other contingent valuation studies of the economic value of ground water quality
have recently been published. The valuation results from these previous studies are summarized in Table 4.
As shown by Table 4, annual willingness to pay for ground water quality from these previous studies ranges
from $129-$1,437 per year. The estimate of willingness to pay obtained in the Dougherty County case study
($641 per year) falls somewhat below the midpoint of this range.
Bergstrom and Boyle18 provide a more detailed comparison of the results reported in
Table 4. In general, they suggest that the variations in these results are attributable largely to differences in:
1) the quality change being valued (e.g., type of contaminant), 2) supply uncertainty, and 3) socioeconomic
variables (e.g., age, income). Some of the observed variations may also be attributable to differences in
questionnaire design (e.g., information, bid elicitation procedure) and estimation techniques (e.g., functional
form). The wide variation in values reported in Table 4 are indicative of the complexity of valuing ground
water quality changes.
GW2-11
-------
Session GW-2
Bergstrom
Table I
Descriptive Statistics for Independent Variables
Variable
M
OWN
SUPP
DEMN
AGE
Mean
42517
3.43
0.541
0.675
46.8
Minimum
5000
1
0
0
19
Maximum
500000
4
1.00
1.00
83
Standard
Deviation
35490
0.790
0.281
0.318
15.7
Table 2
Logit Analysis of Ground Water Protection
Variable
Constant
log(M)
log(OWN)
log(SUPP)
log(DEMN)
log(AGE)
Coefficient
-1.08
0.7378
1.49*
0.363a
0.0732
-0.718b
McFadden Rc 0.267
Number of Obs. 591
alndicates significance at 1% level.
blndicates significance at 5% level.
°See Note b one page GW2-5.
GW2-12
-------
Session GW-2
Bergstrom
Table 3
WTP Sensitivity Analysis
Variable
M
OWN
SUPP
DEMN
AGE
Value
7,027b
78,007°
1 (min.)
4 (max.)
0 (min.)
1.00 (max.)
0 (min.)
1.00 (max.)
31.1b
62.5C
Option Price ($/yr)*
165
1452
71
905
120
942
451
682
870
469
aThe estimated option price using means of the variable has a mean of $641/yr. and a median of $636/yr.
blndicates that the value is one standard deviation below the mean value.
Indicates that the value is one standard deviation above the mean value.
Table 4
Comparison of CVM Ground Water Valuation Study Results
Study Area
Dougherty County, GA
Dover, NH
Cape Cod, MA
Valuation Issue
Protection of potable water from
contamination3
Protection of potable water from
contamination3
Protection of potable water from
nitrate contamination
WTR
$641/year/household
(CI: $493-$890/year/household)
$129/year /household
$363-$l,437/year/
household
"Type of contaminant not specified
Adapted from: Bergstrom and Boyle, 199218
GW2-13
-------
Session GW-2
Bergstrom
7.0 IMPLICATIONS
The results of this case study suggest that a select group of citizens in Dougherty County
highly value the protection of ground water from uncertain future contamination. It is important to note that
these results must be interpreted with caution. The valuation results directly apply to only a specific small
region - Dougherty County, Georgia. The sample of Dougherty County residents was also rather limited in
that only registered voters were sampled. Registered voters may not accurately represent the general
population hi terms of variables such as mean income, mean age, and pollution control preferences.
Although limited, the results provide some useful insights into the value of ground water
protection programs. A general question of potential interest to agencies such as the U.S. EPA is: "Will the
public support programs for protecting ground water from potential agricultural chemical contamination,
even if these programs result in higher costs which reduce annual income levels?". The Dougherty County
case study suggests that the voting public is likely to support ground water protection programs, at least in
areas like southwest Georgia where much of the public is dependent on ground water for drinking water and
there is a real threat of future contamination.
The economic value of a ground water protection program is likely to vary across space and
time because of differences in variables that influence an individual's willingness to pay (or willingness to
accept compensation) associated with the program. In the Dougherty County case study, variables which
appeared to significantly influence an individual's willingness to pay for ground water protection included
supply uncertainty (e.g., probability of future contamination without a protection program), socioeconomic
variables (e.g., annual income, age), and health preferences (e.g., an individual's concern about the effects of
environmental pollution on his or her own health). These results suggest that variations in these variables
across different regions of the U.S. may increase or decrease the benefits of ground water protection
programs. For a given region of the country, such as Dougherty County, the results also suggest that
changes in these variables over time may increase or decrease the benefits of ground water protection
programs. Changes in the perceived benefits of ground water protection programs, in turn, would be
expected to influence the actions of individuals with respect to supporting or not supporting ground water
protection (e.g., voting behavior, defensive expenditures).
The results of the Dougherty County case study also have implications with respect to future
ground water valuation research. First, the contingent valuation method appeared to provide reasonable
estimates of the benefits of protecting ground water from agricultural chemical contamination. The valuation
responses, for example, were strongly correlated with explanatory variables suggested by economic theory.
There is also evidence of convergent validity when the results of the Dougherty County study are compared
with the results of comparable previous studies (Edwards's19 study in particular). Further testing of the
contingent valuation method as a means for valuing the benefits of ground water quality protection therefore
seems warranted.
Future applications of the contingent valuation method to value ground water quality
protection should consider several issues. It is very important in any contingent valuation application to
carefully define the "commodity1 being valued. In the case of a ground water quality protection program, a
researcher needs to carefully consider the "with program" and "without program" scenarios so that the change
in ground water quality provided by the program can be properly delineated. The next challenge is to clearly
communicate this change to respondents through the judicious presentation of information hi the contingent
valuation questionnaire.
Another important issue in need of more research is the determination of the relevant
population to sample in a particular contingent valuation study. The general question to be answered is "For
this particular ground water quality valuation problem, whose opinions and valuations matter?" Answering
this question involves defining the "market area" for the services provided by the ground water resource of
interest. Services provided by a ground water resource may include use values and indirect use values (e.g.,
GW2-14
-------
Session GW-2
Bergstrom
bequest values). Definition of the "market area" for a ground water resource is complicated by the need to
decide the extent to which the various direct use and indirect use values will be accounted for in a particular
ground water valuation study.
Future research also needs to be devoted to determining how to more effectively deal with
"protest bidders" when analyzing contingent valuation data. In most contingent valuation studies, "protest
bidders" are screened out of the final valuation data set. These "protest bidders", however, may provide
valuable information for public policy purposes. For instance, "protest bidders" may represent a segment of
the general public who would strongly oppose an actual, proposed ground water protection program which
would end up costing them something in terms of reduced annual income.
A potentially useful product of a coordinated ground water research program would be
development of a standard protocol for applying the contingent valuation method to value ground water
quality protection or restoration. Development of such a standard protocol would include considerations
related to survey questionnaire design, sample selection, survey implementation, data analysis, and the
interpretation and application of results.
2.0 REFERENCES
1. Nielsen, E.G. and L.K. Lee. "The Magnitude and Costs of Ground Water Contamination
from Agricultural Chemicals." Agricultural Economic Report. No. 576. ERS/USDA. 1987.
2. Williams, W.M., P.W. Holden, D.W. Parsons, and M.N. Lorber. Pesticides hi Ground
Water Data Base: Interim Report. U.S. EPA, Office of Pesticide Programs, Washington,
D.C. 1988.
3. U.S. Environmental Protection Agency. Agricultural Chemicals in Ground Water: Proposed
Pesticide Strategy. U.S. EPA, Office of Pesticides and Toxic Substances, Washington, D.C.
1987.
4. Rouhani, S. and TJ. Hall. "Geostatistical Schemes for Ground Water Sampling."
J. Hydrology. Vol. 103, pp. 85-102. 1988.
5. Cohen, S.Z., S.M. Creeger, and C.G. Enfield. "Potential Pesticide Contamination of Ground
Water from Agricultural Uses," Treatment and Disposal of Pesticide Wastes. ACS
Symposium Series 259, Washington, D.C.. 1984.
6. Kundell, J.V. (Editor). Georgia Water Resources: Issues and Options. Institute of
Government, The University of Georgia, Athens, GA. 1980.
7. Georgia Department of Natural Resources (DNR). Georgia Nonpoint Source Assessment
Report. Georgia DNR, Environmental Protection Division, Atlanta, GA. 1989.
8. Bishop, Richard C. "Option Value: An Exposition and Extension," Land Economics.
Vol. 58, pp. 1-15. 1982.
9. Smith, V.K. "Option Value: A Conceptual Overview," Southern Economic Journal. Vol. 49,
pp. 654-658. 1983.
10. Cameron, TA. "A New Paradigm for Valuing Non-Market Goods Using Referendum Data:
Maximum Likelihood Estimation by Censored Logistic Regression," J. Envir. Econ. Manag.
Vol. 15, pp. 355-79. 1988.
GW2-15
-------
Session GW-2
Bergstrom
11. Hanemann, M.W. "Welfare Evaluations in Contingent Valuation Experiments with Discrete
Responses," Amer. J. Agri. Econ. Vol. 66, pp. 332-41. 1984.
12. McConnell, K.E. "Models for Referendum Data: The Structure of Discrete Choice Models
for Contingent Valuation," Journal of Environmental Economics and Management. Vol. 18,
pp. 19-34. 1990.
13. Bowker, J.M. and J.R. Stoll, "Use of Dichotomous Choice Nonmarket Methods to Value the
Whooping Crane Resource," Amer. J. Agri. Econ. Vol. 70, pp. 372-81. 1988.
13. Dillman, DA. Mail and Telephone Survey: The Total Design Method. New York. John
Wiley & Sons. 1978.
14. Bergstrom, J.C., J.R. Stoll, J.P. Titre, and V.L. Wright. "Economic Value of Wetlands-
Based Recreation," Ecological Economics. Vol. 2, pp. 129-47. 1990.
15. Cummings, R.G., D.S. Brookshire, and W.D. Schultze, Valuing Environmental Goods: An
Assessment of the Contingent Valuation Method. Totowa, New Jersey. Rowan and
Allanheld. 1986.
16. Desvousges, W.H., V.K. Smith, and A. Fisher. "Option Price Estimates for Water Quality
Improvement: A Contingent Valuation Study for the Monongahela River," J. Envir. Econ.
Manag. Vol. 14, pp. 248-67. 1987.
17. U.S. Department of Commerce. 1990 Census of Population and Housing. Economics and
Statistics Administration, Bureau of the Census, Washington, D.C. 1991.
18. Bergstrom, J.C. and K.J. Boyle. "Benefit Transfer Case Study: Benefits of Ground Water
Protection in Dougherty County, Georgia," Proceedings from the AERE Workshop of
Benefit Transfer (forthcoming). 1992.
19. Edwards, Steven F. "Option Prices for Ground Water Protection," J. Envir. Econ. Manag.
Vol. 15, pp. 475-87. 1988.
GW2-16
-------
Session GW-2
Abdalla
Avoidance Costs and Ground Water Values;
Results of Two Empirical Applications
Charles W. Abdalla
Associate Professor of Agricultural Economics
Department of Agricultural Economics and Rural Sociology
Pennsylvania State University
ABSTRACT
The cost avoidance (or averting expenditure) approach was applied to measure household-
level economic losses resulting from ground water contamination. This approach is grounded in economic
theory and, provided certain assumptions are met, can provide lower bound estimates of the value of
environmental quality improvements. Mail surveys were used to collect data on avoidance actions and
household characteristics in communities in central and southeastern Pennsylvania served by public wells
containing volatile organic chemicals. Households' knowledge of contamination and avoidance expenditure
levels varied significantly between the two study sites. In the central site, 96% of households were aware of
water contamination and 76% of those with such knowledge undertook avoidance behaviors. Only 43% of
households in the southeastern site were aware of contamination. Of those, about 44% undertook avoidance
actions. Costs averaged $5.25 (1987 dollars) and $.40 (1989 dollars) per week for each household that chose
to avoid the contaminant in the central and southeast study sites, respectively. Common factors influencing
the likelihood that a household would undertake avoidance actions included households' qualitative rating of
the ground water contaminant's health risk, the amount of information acquired about the contaminant or its
health risk, and presence of children within the household.
1.0 INTRODUCTION
Findings of contaminants in ground water have been reported with increasing frequency.
However, the human health and economic consequences of ground water contamination are not well known.
Despite these uncertainties, people are demanding that pubh'c policies be changed to better protect ground
water. There is a need for information about the costs and benefits of various policy options for managing
ground water. One objective of the research reported in this paper was to generate information about the
costs of not adequately protecting ground water. A second goal was to explore the potential of the avoidance
cost (averting or defensive expenditure) approach for measuring the benefits of environmental improvements.
A final goal was to increase understanding of the process by which people respond to information about
environmental health risks in the ground water context.
2.0 AVOIDANCE COST APPROACH
The avoidance cost approach is an indirect valuation technique that infers the value of
environmental quality changes from consumption of related goods or services in the economy. Valuation is
based on evidence of actual behaviors of consumers or producers under real budget constraints, rather than a
hypothetical choice framework used in contingent valuation studies. This method has a sound basis in
economic theory yet relatively few empirical applications have been completed.
2.1 Conceptual Framework
The avoidance cost method is rooted in the household production function model of
consumer behavior.1 This model's basic idea is to view consumers as producers. In this framework,
households do not derive utility directly from purchased goods, but rather use marketed and nonmarketed
goods as inputs to produce outputs that have value to them. Observations of household choices regarding
GW2-17
-------
Session GW-2
Abdalla
consumption of inputs which complement or substitute for environmental quality can provide information
which will allow benefit estimation.
Most theoretical treatments of averting expenditures have concluded that these expenditures
provide a lower bound estimate of the true cost of increased pollution. Courant and Porter2 and Bartik3
demonstrated that the savings in averting expenditures, holding the level of environmental quality constant,
are equal to the benefits of a marginal pollution reduction. In reality, individuals will alter their choice of
environmental quality level as pollution is reduced. Consequently, observed changes in averting expenditures
provide a lower bound to the benefits of a marginal reduction in pollution.
The theoretical framework developed by Bartik3 to evaluate nonmarginal changes in
environmental quality provided the conceptual foundation for the research efforts described in this paper.
Bartik demonstrated two alternative measures of the benefits of pollution reduction, the compensating
variation and equivalent variation, can be obtained by estimating the demand for "personal environmental
quality" or Q. Compensating variation and equivalent variation are theoretical refinements of consumer
surplus, the standard measure of welfare change resulting from environmental improvements. Environmental
quality changes can increase or decrease the price of a good to an individual and therefore affect consumer
surplus. In Bartik's model, households choose their personal environmental quality level by adjusting
averting expenditures given an externally determined level of pollution. Since estimating the household
demand for Q can be costly, two theoretical measures were derived which bound the true compensating
variation and equivalent variation measures of the benefits of a pollution decrease. Bartik contends that less
information is needed to estimate the lower and upper bound measures than the exact benefit measures
which require information on the behavioral responses at the household's chosen Q. Only the lower bound
measure is discussed below.
The lower bound to the compensating variation measure of a pollution reduction is given by
the savings in defensive expenditures needed to reach Q0, the level of personal environmental quality chosen
before the pollution reduction. The key assumption is that households maintain their original Q. For
example, if pollution levels fall and households remain at the personal environmental quality level of Q0, then
the household obtains a benefit equal to the savings on averting expenditures. If adjustments in personal
environmental quality are permitted, then the household must be even better off. Under this restriction
regarding Q, averting expenditure savings represent a lower bound to the compensating variation measure of
benefits.
The avoidance cost approach rests on several key assumptions. First, averting expenditures
must be perfect substitutes for pollution reductions. In other words, inputs purchased to avoid contamination
should not provide other value to households. Second, pollution must not directly affect household utility.
Third, averting expenditures should not involve purchase of durable goods.4 If these assumptions do not
hold, averting expenditures savings are no longer an accurate lower bound estimate of environmental
improvements.
22 Application to Ground Water
Ground water contamination can have many consequences for a community, including losses
related to human health effects, avoidance costs incurred by households, businesses and municipalities,
increased fear and anxiety, ecological damages, and reduced nonuser benefits, such as option or existence
values. The studies presented in this paper were designed to empirically measure only household-level
avoidance costs resulting from ground water contamination.
As stated earlier, one research goal was to investigate the potential of the avoidance cost
method for evaluating ground water contamination impacts. In order to apply the method, information on
the quantity and price of averting inputs are needed. Actions that can be taken to avoid exposure to a
ground water contaminant include purchasing bottled water, buying a home treatment device, boiling water
(for some contaminants), hauling water from another source, and changing food or beverage consumption.
GW2-18
-------
Session GW-2
Abdalla
3.0
EMPIRICAL APPLICATIONS
Following Dillman,5 mail questionnaires were used to collect data on averting actions in
communities in central and southeastern Pennsylvania served by public water supply wells containing volatile
organic chemicals. Only residential water customers received questionnaires. Thus, costs to businesses, the
water supplier and government agencies were not addressed. Information was collected about possible
household-level factors that may influence averting expenditures, such as health perceptions, attitudes, and
demographic factors.
3.1
Central Pennsylvania Survey Site
This particular application was a pilot effort to explore how households respond when
ground water contamination occurs. Greater detail on the study can be found in Abdalla6. The survey site
is located in the geographical center of the state in Centre County and includes several communities near
State College, Pennsylvania. In this area, an investor-owned public water supply serves about 1,800
customers or about 5,000 people. The contaminant was perchloroethylene (PCE) and the levels ranged from
20 to 32 parts per billion (ppb). The public notification occurred in July 1987 but records indicated that the
contaminant was detected by authorities in the early 1980s but not made public at that time. No drinking
water safety standard existed for PCE in 1987. Risk information was provided to affected households by the
Pennsylvania Department of Environmental Resources. Individuals made the decision about the water's
safety and use in their household.
The population was served contaminated water for a six-month period following public
notification. The system was connected to an uncontaminated water source in December 1987. The
questionnaire was sent to all residential customers in February 1988. After several follow-ups, a survey
response rate of about 70% was obtained. The survey results were used to estimate adjustments among the
population of about 1,600 households.
Ninety-six percent of the households responding to the questionnaire were aware of PCE
contamination. Of these households, about 76% undertook some averting action (Table 1). Bottled water
purchases were the most frequent averting behavior undertaken (Table 2). Almost 50% of those surveyed
started buying bottled water after public notification. Households that purchased bottled water before public
notification increased their consumption. Almost one-quarter of the households boiled water to reduce PCE
levels while very few people purchased home water-treatment devices.
Table 1
Awareness of Contamination and Avoidance Behaviors Among Samples
of Households in Central and Southeastern Pennsylvania Survey Sites
Percent of Households Aware of
Ground Water Contamination
Percent of Informed Households
Taking Avoidance Behaviors
Central Site
96.0
76.4
Southeastern Site
43.2
43.7
Sources: Abdalla (1990) and Abdalla et al. (1992).
GW2-19
-------
Session GW-2
Abdalla
Table 2
Estimated Avoidance Behaviors among Populations of
Households in Central and Southeastern Pennsylvania Survey Sites
Household Avoidance Behavior
New Bottled Water Purchases
Increased Bottled Water
Purchases by Previous Users
Boiling Water
Hauling Water
Home Water Treatment
Estimated Percent of Population -
Central Site1
47.8
15.2
23.0
29.3
3.3
Estimated Percent of
Population - Southeast Site
4.2
7.2
2.9
7.1
3.9
aColumn totals more than 100% since many households undertook more than one averting action.
Sources: Abdalla (1990) and Abdalla et al. (1992).
Economic losses are broken down in Table 3 by four categories: bottled water, hauling
water, boiling water and home treatment. Approximately 1,100 households taking averting actions over the
six-month period spent an estimated $137,371 to $160,343 (1987 dollars), depending on the wage rate used to
reflect the value of lost leisure time. Costs averaged about $5.25 per week for each household. Households
paid an out-of-pocket cost of $21 to avoid exposure to the contaminant. This was three times their monthly
water bill at that tune of about $7 per month.
Table 3
Estimated Avoidance Costs Incurred by Households at
Central Pennsylvania Survey Site from July to December 1987
Household Averting Action
Bottled Water Purchases
Hauling Water
Boiling Water
Home Water Treatment
Total
Low Estimate*
(1987 Dollars)
$71,876.84
$38,227.39
$12,328.93
$14,938.04
$137,371.20
High Estimate1*
(1987 Dollars)
$71,876.84
$45,542.89
$27,985.64
$14,938.04
$160,343.41
aLeisure time was valued at 1987 minimum wage of $3.35 per hour.
bLeisure time was valued at average wage for manufacturing in Pennsylvania of $8.27 per hour in 1987.
Source: Abdalla (1990).
GW2-20
-------
Session GW-2
Abdalla
32 Southeast Pennsylvania Survey Site
The second community selected was the borough of Perkasie. This community had an
estimated 2,760 households and is located in Bucks County in southeastern Pennsylvania. Funding for this
research was provided by the U.S. Environmental Protection Agency (EPA). Additional information on this
study can be found in Abdalla et. al.7 and Abdalla et al.8
Trichloroethylene (TCE) was detected in Perkasie's wells in late 1987. TCE levels were as
high as 35 ppb, exceeding the EPA's maximum contaminant level (MCL) of 5 ppb. Since no temporary
solution was available to reduce the levels below the MCL, the borough was required in June 1988 to notify
customers of the contamination. The contaminant was still present when the households were surveyed in
September 1989.
A representative sample of 1,733 was chosen for Perkasie. After three follow-up mailings,
761 usable questionnaires were received providing an effective response rate of 46.9%. Since this rate was
lower than anticipated, telephone interviews with a random sample of 50 nonrespondents were conducted to
determine if respondents and nonrespondents were similar in key attributes. The two samples were not
found to be significantly different and the survey results were concluded to be representative of the
population.
Only 43.2% of households in the southeast site were aware of contamination. Of those,
43.7% undertook avoidance actions (Table 1). The avoidance behaviors are broken down by type in Table 2.
The costs of these actions were calculated for the total population of Perkasie residents and are presented in
Table 4. Total losses from December 1987, when TCE was first detected, to September 1989 ranged from
$61,313 to $131,334 (1989 dollars), depending on the wage rate used to reflect the value of lost leisure time.
The average increase in expenditures per household which undertook averting actions in response to the
contamination was $1.60 per month or $0.40 per week.
Table 4
Estimated Avoidance Costs Incurred by Households at
Southeast Pennsylvania Survey Site from December 1987 to September 1989
Household Averting Action
Bottled Water Purchases
Hauling Water
Boiling Water
Home Water Treatment
Total
Low Estimate*
(1989 Dollars)
$28,476.49
$12,512.76
$15,632.58
$4,691.46
$61,313.29
High Estimate1*
(1989 Dollars)
$28,476.49
$34,031.48
$64,134.63
$4,691.46
$131,334.06
aLeisure time was valued at 1989 minimum wage of $3.35 per hour.
bLeisure time was valued at an estimated hourly average wage calculated from the before-tax income
category checked by the respondent in the questionnaire.
Source: Abdalla et al. (1992).
GW2-21
-------
Session GW-2
Abdalla
33 Factors Influencing Avoidance Expenditures
Logistical regression analysis was used to identify factors which influenced avoidance actions.
At the central Pennsylvania site, the most significant predictors of the probability that a household would
undertake avoidance actions were: the qualitative rating of PCE health risks; whether the household
consulted a third party about water use; the amount of information obtained on PCE health risks; the
amount of information obtained on avoidance practices; whether the respondent received periodic medical
examinations; whether a pregnant woman was present in the household; whether children under the age of 5
were present in the household; and trust in state and local institutions. A positive relationship was found to
exist between the dependent variable representing the decision to avert and the independent variables with
the exception of the amount of information obtained on PCE risk and the trust variables. Consequently, the
likelihood that a household would undertake an averting action decreased as households obtained more
information on the health risks of PCE and as their trust in state and local institutions increased.9 Logistical
regression analysis of the southeast data indicated that households were more likely to take averting actions
if they received information about TCE, rated the cancer risks associated with TCE levels hi their water to
be relatively high, or if children ages 3 to 17 were present. Analysis using ordinary least squares techniques
found that the presence of children less than 3 years old in a household was associated with increased
intensity of averting actions.7
3.4 Observations about the Results
Households' knowledge of contamination, averting responses, and total expenditure levels
varied greatly between the two study sites. The differences in averting expenditures are probably related to
differences in the local situations and characteristics of the risk. In the central site, 96% of households were
aware of water contamination and 76% of those with such knowledge undertook avoidance behaviors (Table
1). Costs averaged $5.25 (1987 dollars) per household per week. In this situation, the media provided
extensive coverage of the contamination episode and considerable public discussion occurred. Uncertainty
about the contaminant's risk was great since local experts disagreed about the risk of the water in the
absence of a drinking water standard. Also, the public water supplier had credibility problems with some
households as a result of previous service problems.
Only about 43% of households in the southeastern site were aware of contamination (Table
1). Of those, almost 44% undertook avoidance actions and costs averaged $.40 (1989 dollars) per week for
each household that chose to avoid exposure. A drinking water standard existed for this contaminant,
reducing uncertainty and perhaps controversy. However, the more important factors are likely to have been
the relatively ineffective public notification efforts of local agencies and the water supplier and lack of
significant media coverage.
Several common factors were found to influence the likelihood that a household would
undertake avoidance actions. These included households' qualitative rating of the ground water
contaminant's health risk, the amount of information acquired about the contaminant or its health risk, and
presence of children in the household.
4.0 CONCLUSIONS
Based on these two empirical applications, it appears that the avoidance cost approach has
considerable potential to estimate certain ground water values. The method has sound theoretical
underpinnings and valuation is based on evidence of actual choice. However, the values obtained are
incomplete in two ways: they represent lower-bound estimates of benefits and do not capture other damages
associated with ground water contamination, such as losses from actual health effects, increased fear and
anxiety, ecological damages, and reduced non-user benefits. Averting losses incurred by businesses and
municipalities were also not investigated in these two studies. While the avoidance cost approach does not
encompass all impacts, it does yield theoretically supported estimates of an important category of the ground
GW2-22
-------
Session GW-2
Abdalla
water values. Public officials should seek information about other values to build upon the base that cost
avoidance studies can provide.
At least two promising research opportunities are evident. First, the validity of the
assumptions underlying the averting cost method deserve further exploration. Key assumptions are: averting
inputs are perfect substitutes for pollution; pollution should not be directly valued by households; and
averting expenditures should not involve the purchase of durables. Future studies should attempt to verify
the extent to which the assumptions do not hold in the context of ground water contamination situations.
For example, the substitutability assumption could be examined by asking individuals about the expected
benefits of averting actions.
The second promising area of future research lies in coupling the avoidance cost approach
with other valuation methods (Abdalla et al. 1990). As noted earlier, the avoidance cost method is capable
of generating lower bound estimates of ground water values and these values have their basis in actual
household decisions made under a budget constraint. These values could be used to serve as an "anchor" in
a choice framework in a contingent valuation study. For example, after avoidance expenditures were elicited
in a mail questionnaire or interview, survey respondents could be asked questions about additional values
they may have for use or nonuse aspects of the ground water resource.
5.0 REFERENCES
1. Becker, G.S. "A Theory of the Allocation of Time," The Economic Journal. Vol. 75,
pp. 493-517. 1965.
2. Courant, P.N. and R.C. Porter. "Averting Expenditures and the Costs of Pollution," Journal
of Environmental Economics and Management. Vol. 8, No. 4, pp. 321-329. 1981.
3. Bartik, T.J. "Evaluating the Benefits of Non-Marginal Reductions in Pollution Using
Information on Defensive Expenditures," Journal of Environmental Economics and
Management. Vol. 15, pp. 111-127. 1988.
4. Dickie, M. and S. Gerking. Benefits of Reduced Soiling from Air Pollution Control: A
Survey. Migration and Labor Market Efficiency. Amsterdam. Martin Nijhoff. 1988.
5. Dillman, DA. Mail and Telephone Surveys: The Total Design Method. New York. John
Wiley, 1978.
6. Abdalla, C.W. "Measuring Economics Losses from Ground Water Contamination: An
Investigation of Household Avoidance Costs," Water Resources Bulletin. Vol. 26, No. 3,
pp. 451-463. 1990.
7. Abdalla, Charles W., BA. Roach, and DJ. Epp. "Valuing Environmental Quality Changes
Using Averting Expenditures: An Application to Ground Water Contamination." Land
Economics. Vol. 68, No. 2, pp. 163-169. 1992.
8. Abdalla, C.W., D J. Epp, and BA. Roach. Developing Methods to Measure Costs of
Averting Behavior in Evaluating Water Quality Improvements. Final Project Completion
Report to US EPA, Office of Policy, Planning and Evaluation. 1990.
9. Abdalla, C.W. and A.G. Rodriguez. "About the Existence of Averting Behavior: A Case
Study in Central Pennsylvania," Department of Agricultural Economics and Rural Sociology,
Pennsylvania State University, University Park, Pennsylvania, unpublished manuscript. 1991.
GW2-23
-------
Session GW-2
Lichtenberg
Using Risk Assessment in Analyzing Ground Water Protection Policies
Erik Lichtenberg
Associate Professor
Department of Agricultural and Resource Economics
University of Maryland
ABSTRACT
Human health risks from contaminants are a major motivation for protecting ground water
quality. Analyzing policies for ground water protection requires evaluating tradeoffs between these health
risks and the cost of diverting resources from other sectors of the economy, which, in turn, requires
quantitative estimation of health risks. Current methods of quantitative health risk estimation introduce a
degree of error, or uncertainty, large enough that it must be taken into account in making decisions. A
practical way of doing this is to combine probabilistic risk assessments with a safety-fixed decision criterion in
deriving uncertainty-adjusted cost curves that relate cost simultaneously to the risk standard and the
probability of a violation. This approach is illustrated using the case of drinking well water contamination in
California by the pesticide DBCP. One important finding is that the marginal cost of health risk reduction
falls substantially as the probability of a violation is made smaller, suggesting that decisions made on the
basis of average risk can lead to underprotection of human health. These uncertainty-adjusted cost curves
can be used in standard cost-benefit analyses, provided that the probability of violation is taken into account,
either by specifying a level that corresponds to standard scientific notions or by estimating individuals'
willingness to pay for different probabilities of violation.
1.0 INTRODUCTION
One of the many reasons for protecting ground water quality is that many chemicals that
leach into ground water pose serious threats to human health. Examples include nitrate, the most commonly
found ground water contaminant in the U.S. today, which causes methemoglobinemia in bottle-fed infants
and has been linked with gastric caner; volatile organic compounds and pesticides found in ground water that
are suspected of causing long-term health problems such as cancer and birth defects; and contaminants like
the insecticide aldicarb that can cause acute, sometimes fatal, poisonings.
Decisions about ground water protection policies must thus take into account tradeoffs
between increased risks to human health and safety, on the one hand, and the costs of diverting resources
from other sectors of the economy, on the other. Economists have considerable experience in estimating the
cost of such policies. The issue of valuing reductions in health and safety risks has also received a great deal
of conceptual and empirical attention (see for example1'2). But economists have paid only limited attention
to the way that heath risks are estimated and the implications of health risk estimation procedures for policy
analysis. The prevailing attitude is one of complete division of labor: let the risk analysis experts take care
of estimating risks, and economists will do their part by assignment of appropriate values to risk reductions.
I have argued previously that certain key factors of quantitative risk assessment as currently
practiced have important implications for policy analysis. '4 This paper discusses the ways in which health
risks are estimated and draws out some lessons for how to incorporate risk estimates into policy analysis. I
then present an example based on previous work on ground water contamination.5
GW2-24
-------
Session GW-2
Lichtenberg
2.0 RISK ASSESSMENT AND UNCERTAINTY
Human health risks from chemicals in ground water are a combination of two factors,
exposure and potency, both of which must be estimated. Exposure is determined by geochemical and
behavioral factors, including leaching of chemicals into aquifers, degradation processes taking place as
leachates move downward into and through aquifers, dilution due to mixing of water in aquifers, and drinking
water ingestion rates. Potency is determined by a host of biological factors.
Quantitative risk assessment employs process models that mimic the chain of events leading
to the creation of a human health risk. Each stage is characterized by a functional form and set of
parameters. The parameters are estimated using either direct or indirect data, and then submodels for each
stage are combined into an overall estimate of risk. For example, one common specification is to express a
health risk, say, the risk of cancer, R, that an individual incurs from ingesting well water containing pesticide
residues over a lifetime, as a multiplicative combination of parameters describing such factors as pesticide
use in the recharge zone, Xj, the concentration of the pesticide leached into well water per unit of pesticide
used, X2, the average ingestion rate for drinking water, X3, the breakdown of the residue into toxic
metabolites, X4, and the toxicity of the pesticide and its metabolites, X5:
R = X1X2X3X4X5 (4)
A quantitative risk assessment would estimate the parameters {X1; X2, X3, X4, X5} using direct information
when available and indirect information otherwise. For example, average use of the pesticide in the recharge
zone might be used to estimate Xj. Either a combination of data on pesticide use and associated pesticide
concentrations hi well water or simulation models like CREAMS might be used to estimate X2. Studies of
human diets might be used to estimate X3. Information on the chemistry of the pesticide and human
metabolism might be used to estimate X4. Because epidemiological data is almost never available, toxicity
data (estimates of X5) are typically derived from rodent bioassays.
Quantitative risk estimates are characterized by substantial uncertainties, because of
variability in geochemical, biological, and behavioral conditions, because of data limitations and because of
our limited knowledge about the fundamental processes involved.
All of the factors involved in the process of creating risk, be they geochemical, biological, or
behavioral, are subject to substantial variability. Leaching rates are highly dependent on rainfall (most
leaching occurs during large storms). Degradation rates depend on water movement, soil composition,
biological conditions within soils, and so on, all of which vary significantly across time and space. Human
beings differ in their drinking water ingestion rates, metabolism, and general susceptibility to toxic agents.
In principle, these variabilities could be captured empirically given sufficient data. But time
and cost militate against a data-intensive solution to this problem. Conditions vary substantially from case to
case, so a generic database is of limited help; this is especially true of ground water because of the wide
variations in hydrogeological conditions. Policy decisions must be made in a timely fashion, and the budgets
of regulatory agencies simply do not permit massive expenditures on regulatory analyses.3
Furthermore, our knowledge about the underlying processes involved is insufficient to
permit theory to substitute for data. For example, we know relatively little about precisely how low-level
exposures to toxicants induce long-term health problems like cancer or birth defects. Theoretical knowledge
gives little firm guidance as to the appropriate functional form of a dose-response relationship or the
reasonable range for potency parameters. The controversial nature of cancer potency estimation is a
"More generally, the issue of how much to spend on reducing uncertainty is an interesting and important
economic question.
GW2-25
-------
Session GW-2
Lichtenberg
reflection of this situation. To a large extent, the same can be said for leaching processes, for chemical
changes in leached contaminants in soils and subsoils, and for movement of contaminants in aquifers.
The combination of unmeasured variability, scientific uncertainty, and plain randomness
suggests that health risk estimates are best viewed as stochastic, that is, that the risk assessment procedure
produces a set of risk estimates, R, characterized by a probability distribution, F(R).
3.0 UNCERTAINTY AND POLICY ANALYSIS: THE STATUS QUO
Uncertainty is not only present in these problems, it is an important factor in decision
making. The evidence suggests that these uncertainties affect both the public's perception of risk and then-
demand for policy intervention. Psychological studies indicate that the public associates greater hazard with
situations that have greater uncertainty associated with them (for a summary see6). The apparent
widespread belief that pesticide residues on foods (e.g., Alar on apples) are a serious problem also bears this
notion out, since the best data available suggest that roughly 60 to 65% of foods in the marketplace have no
detectable residues and almost all of the remaining cases involve residue levels that are extremely small and
well below those the EPA considers the maximum safe levels. A similar situation applies to pesticides in
ground water. While the public believes that ground water pollution from pesticides is a serious problem, a
recent EPA survey found that only 1% of drinking water wells contained pesticide residues in excess of a
health-based standard.8b Policy makers also appear to be quite sensitive to these uncertainties, perhaps
because of the public's sensitivity to uncertainty and perhaps because of an error-avoidance mentality, since
mistakes are a highly visible indicator of poor performance.
The importance of uncertainty makes risk estimates essentially multidimensional. A risk
estimate, R, cannot be considered a scalar quantity, because parameters characterizing uncertainty about the
estimates (i.e., its probability distribution, F(R)) must also be taken into account. This has two implications
for economic, or, more broadly, policy analysis. First, uncertainty must be incorporated into the decision
methodologies used for policy analysis: both benefits and costs must be adjusted explicitly for uncertainty.
Second, this adjustment must be handled in a consistent manner so that decisions are comparable from one
situation to another.
The current prevailing approach meets neither of these criteria. Instead, economists have
tended to ignore the complications raised by uncertainty. Uncertainty is left to risk analysts to deal with
during the court of risk assessment. Economists then take the risk estimates as certain (or certainty
equivalents) and use them in an unmodified cost-benefit framework, combining them with estimates of
average values of risk reduction (life-saving) and average costs under alternative policies. Thus, the task for
economists is largely accounting, (i.e., pricing risk reductions and costing out potential solutions).
This approach has two serious shortcomings. First, the procedure used to adjust for
uncertainty produces estimates that are idiosyncratic in each case and are thus qualitatively noncomparable.
Second, it ignores the fact that different policies affect the entire distribution of estimated risk F(R) in
different ways, and thus eliminates consideration of tradeoffs between reducing risk on average and reducing
uncertainty about risk.
Consider first the problem of uncertainty adjustment in risk assessment. The response of
EPA on other regulatory agencies to sensitivity about uncertainty has been to incorporate uncertainty
adjustments into quantitative risk assessments: essentially, to overestimate risk. Specifically, each parameter
estimate is taken from a worst-case scenario. Functional forms are also chosen to be conservative, that is, to
overestimate risk. The resulting risk estimates are thus adjusted upward as a hedge against uncertainty. This
bCurrent ground water quality conditions do not fully measure the scope of this problem because of the long lag
time involved in leaching (easily 30 years in some cases) and because of the rapid growth of pesticide use in the
past. Thus, the current low level of detections may be a harbinger of a much larger problem in the future.
GW2-26
-------
Session GW-2
Lichtenberg
bias in uncertainty adjustment comes from the preventive posture characteristic of the public health
profession, which weights avoiding false negatives more heavily than avoiding false positives. (Much of the
enabling legislation takes the same point of view, as can be seen from their requirements that public health
be protected with an adequate margin of safety.)
The risk model (1) can be used to illustrate this uncertainty adjustment procedure. Suppose
that each of the parameters of that model {X1( X2, X3, X4, X5] is distributed lognormally, so that the log of
the estimated risk, r, is distributed normally. It will be convenient to work with the log of risk, which,
because the transformation is monotonic, changes nothing essential. Let Xj denote the log of the ith
parameter, /ij, its mean and a?, its variance. To give the notion of worst-case scenario a formal statistical
basis, suppose that the parameter associated with such a scenario is the upper limit of a confidence interval
with confidence level o; for example, a would usually be 95 to 99%. The risk estimate derived in this way is:
(5)
where Z(a) is the value of the standard normal distribution exceeded with probability 1-a. The risk estimate
exceeded with probability a is:
5
E Mi + Z(a)
(6)
It should be apparent that r(a ') is the upper limit of a confidence interval with a confidence level a/ greater
than a. Moreover, the confidence level associated with each risk estimate of the type r(a') will vary from
case to case, making risk estimates noncomparable from one situation to another. This is disturbing because
it undermines the basis for applying cost-benefit criteria.0
Next consider the issue of modeling policy-induced changes in the risk distribution, F(R).
The standard cost-benefit procedure used makes no explicit assumptions about these effects. Instead, it
models policy options using percentage reductions in different parameters. In the lognormal model
considered here, this translates into an implicit assumption that the mean and variance of the log of each
parameter are reduced by the same percentage. For example, assuming that a policy reduces parameter i by
a fraction ^j implies that the mean becomes (1-0 ;) ^j and the variance (1-^j)2 a\, and that the log risk
becomes:
i-l M
This assumption is not particularly realistic: one would expect different policy options to
have different effects on the mean and variance of the estimated parameters. For example, collecting
additional information is clearly a policy option; it has no effect on the expected value of the risk estimate,
but should increase its precision (reduce uncertainty). Reducing pesticide applications rates may lower
expected leaching into ground water without altering uncertainty about leaching, or it may change expected
leaching and uncertainty about leaching in different ways. In general, one would expect policy decisions to
cln practice, the situation is worse because worst-case scenarios are not selected according to specific
statistical criteria. As a result, uncertainty adjustments in risk estimates vary unpredictably, making it
difficult to apply economic criteria to issues such as targeting risk reduction efforts.
GW2-27
-------
Session GW-2
Lichtenberg
involve tradeoffs between reducing risk on average and reducing uncertainty about risk; the standard cost-
benefit methodology currently used is unable to analyze such tradeoffs because it assumes away their
existence.
4.0 AN ALTERNATIVE APPROACH
The implication of the preceding analysis is that risk assessment cannot remain a black box
to economists (or, more broadly, policy analysts). Policy analysts must work with risk assessments that
incorporate explicitly: (a) statistically based adjustments for uncertainty, and (b) model the different effects
of policy on both average risk and uncertainty about risk. In other words, economists must work with
probabilistic risk assessments that produce estimates of the probability distribution of the risk estimate F(R)
and they must explicitly model the effects of alternative policies on that entire distribution.
That leaves the issue of the appropriate decision framework to use for choice under
uncertainty. Because of the public's sensitivity to uncertainty, cost-benefit analysis using expected risks and
costs does not seem appropriate. The standard alternative is expected utility, implemented using methods
such as multiattribute decision analysis or revealed preference estimation of policy preference parameters.
On a practical level, these approaches are problematic in a social choice context. Multiattribute decision
analysis assumes that the preferences of a single decision-maker completely mirror society's. Revealed
preference methods assume that past decisions completely reflect current policy preference parameters,
despite the fact that political climates clearly change over time. Thus, neither empirical implementation of
expected utility theory seems particularly reasonable.
Expected utility theory is problematic on a conceptual level as well. The theory assumes
that people have preferences defined over outcomes and that the probabilities of these outcomes are known
with certainty, either objectively or subjectively. Assuming that the probabilities of the outcomes are known
only stochastically and that uncertainty about them matters to a decision maker violates the essential
structure of expected utility and necessitates the use of alternative theory of choice under uncertainty.9
Possibilities include Bayesian approaches or more general theories of choice under uncertainty developed to
explain observed anomalies in preferences.
As an alternative, Lichtenberg and Zilberman proposed an approach based on classical
statistical methods for adjusting for uncertainty, specifically, a safety-fixed criterion. Let E(J8) denote the
level of social expenditure associated with a policy vector ft . Their method involves choosing ft to:
min E (ft ) sJ. Pr{R(ft) a , (8)
where a indicates the confidence level desired. The solution to this problem is a policy vector /?*(R0pr) that
is implicitly a function of the estimated risk R0 and the confidence level a and thus a dual uncertainty-
adjusted social cost function:
C(R0p) = {min E(p) si. Pr{R(ft) < *,)}><*/}. (9)
Using a simplified version of the lognormal model defined by equation (3), Lichtenberg and Zilberman show
that this uncertainty-adjusted cost curve is decreasing in risk (9 C/9 R0 < 0) and increasing in the confidence
level (9 C/da > 0), as one would expect. The increase in cost per one percentage point increase in the
confidence level 9 C/da can be interpreted as a premium paid for uncertainty reduction, akin to the risk
premium of standard decision theory of choice under uncertainty. Lichtenberg and Zilberman also show that
the marginal cost of risk reduction, 9 C/9 RQ, is decreasing in risk and in the confidence level.
Lichtenberg and Zilberman note that this approach is attractive because it corresponds to
the legislative mandates of regulatory agencies, in that the confidence level a quantifies the margin of safety
with which public health is to be protected. Moreover, it takes a classical statistical approach that natural
GW2-28
-------
Session GW-2
Lichtenberg
scientists (including risk analysts) are more comfortable with, making it more amenable to interdisciplinary
cooperation.
Once estimated, the uncertainty-adjusted cost curve C(R0pr) can be used in conjunction with
standard tools of decision analysis such as cost-benefit analysis. For example, one could find the appropriate
risk standard (and thus, implicitly, the optimal policy mix) by equating the marginal cost of risk reduction
with willingness to pay for risk reduction. Doing this requires fixing the confidence level a. Lichtenberg and
Zilberman suggest setting a to correspond with the standard of scientific reliability such as 95 or 99%.
Alternatively, a could be set endogenously according to decision-theoretic criteria; I return to this issue
below.
5.0 CASE STUDY: GROUND WATER CONTAMINATION BY A PESTICIDE
Lichtenberg, Zilberman, and Bogen applied this approach to estimating the uncertainty-
adjusted cost of risk reduction to a case involving residues of the nematicide l,2-dibromo-3-chloropropane
(DBCP) found in drinking water wells in Fresno County, California. DBCP, which had been used as a soil
fumigant for orchard crops (primarily peaches), was banned for all agricultural uses by the U.S.
Environmental Protection Agency in 1979 after having been linked to sterility in male chemical plant
operators and gastric cancer hi mice and rats. Because DBCP was no longer hi use, the study ignored the
question of linkages between pesticide use, leaching and consumer and producer welfare in the orchard crop
market. Instead, it focused on tradeoffs between excess gastric cancer risk and the cost of developing clean
drinking water supplies.
A probabilistic quantitative risk estimate of the excess cancer risk faced by an individual
drawn at random from the population of the county was constructed using Monte Carlo simulation. A
multiplicative risk model like equation (1) was used. The parameters were: (a) the concentration of DBCP
in drinking water, (b) the error involved in measuring that concentration, (c) consumption of water by an
individual over a 30-year lifetime, (d) an interspecies dose equivalence factor that converts an animal dose to
a human equivalent, and (e) a carcinogenic potency parameter.
California State Department of Health Services data were used to construct the distribution
of DBCP concentrations in well-based water systems and to estimate the error in measuring DBCP
concentrations. For the sake of illustration, it was assumed that the observed variability in DBCP
concentrations across water systems was known only stochastically. In other words, the analysis treated the
entire area as a single water system with random distribution of DBCP contamination levels. Measurement
error was assumed to be distributed normally for water systems with DBCP concentrations above the
detection limit for DBCP and uniformly for wells with DBCP concentrations below the detection limit.
The distribution of lifetime water consumption was estimated using a triangular density
ranging between 80 and 120% of the mean intake rates for males and females estimated by the International
Commission of Radiological Protection.10 Census data indicated that 49% of the population of Fresno
County was male.
There are two main hypotheses about how to translate animal doses into human equivalents.
One posits calibrating dose on the basis of surface area, the other, on the basis of body weight. These two
hypotheses were assumed to be equally likely to be correct.
The carcinogenic potency parameter and the distribution of this estimate were derived by
applying maximum likelihood to a multistage dose-response model using data from a feeding study of mice.
The costs of developing new, clean water supplies differed between rural and urban areas.
High fixed costs coupled with larger capacity made drilling new wells the least-cost alternative for large
systems, at least in the short to medium run, while installing filtration devices was cheaper for individual
GW2-29
-------
Session GW-2
Lichtenberg
The analysis focused on prioritizing clean water supply development across the 66 districts in
the county. The optimization problem involved minimizing the cost of meeting a risk standard R,, with a
given confidence level a for an individual drawn at random from the county population. It turned out to be
optimal to develop clean water supplies in reverse order of the per capita cost of contamination reduction
(i.e., to clean up the water supply hi the districts with the lowest per capita costs of contamination reduction
first, continuing until the standard defined with the specified confidence level was met). The optimization
problem was repeated over the entire feasible range of standards, resulting in a relationship between total
cost and a distribution of risk (i.e., an uncertainty-adjusted cost curve for risk reduction). To facilitate
analysis, this relationship was smoothed using a second-order polynomial regression of cost on the natural
logarithms of the risk standard and confidence level.
Figure 1 shows the smoothed uncertainty-adjusted total cost curves for risk reduction for
average risk and risk standards with 95 to 99% confidence levels. It is clear that the uncertainty premium in
this case (the additional cost entailed by increasing the confidence level) was large, and that it varied little
Total Costs of Risk Reduction Under
Alternative Confidence Levels
Million Dollars
TCRR - Mean Risk
TCRR-95% CL
TCRR-99%CL
4.000E-06
4.000E-05
Individual Excess Cancer Risk
Figure 1
across risk levels. A one-percentage-point increase in the confidence level increased the total cost of risk
reduction on the order of $3 to 4 million, or between 2 to 10% of the total.
Figure 2 shows the corresponding smoothed uncertainty-adjusted marginal cost curves for
risk reductions of 1 in 10,000. All were decreasing hi risk and in the confidence level, as expected. Marginal
cost was sharply decreasing in risk, ranging between roughly $100 for risk levels on the order of 4 hi 100,000
GW2-30
-------
Session GW-2
Lichtenberg
1200
Marginal Costs of Risk Reduction Under
Alternative Confidence Levels
Dollars per 1/10,000 Reduction in Risk
MCRR - Mean Risk
MCRR - 95% CL
-#- MCRR-99% CL
4.000E-06
4.000E-05
Individual Excess Cancer Risk
Figure 2
to $900 to $1,150 for risk levels on the order of 4 in l,000,000.d The marginal cost of risk reduction was
noticeably lower for higher confidence levels, especially for stringent risk standards (low risks); there were
only slight differences in marginal cost across confidence levels for lax risk standards (higher risks). For
stringent risk standards, the marginal cost of reducing risk on average was 21 to 26% higher than the
marginal cost with a 95% confidence level and 23 to 29% higher that the marginal cost with a 99%
confidence level. Making allowance for the uncertainty can thus reduce the marginal cost of risk reduction
substantially.
This result opens up an interesting perspective on the economic literature critiquing health
and safety regulation in the U.S. It has been widely argued that many health and safety regulations are
overly stringent because the estimated marginal costs of risk reduction are considerably higher than the
marginal benefits implied by benefits estimates derived from studies of the labor market or consumer
behavior.11'12 Once adjustments for uncertainty are taken into account, however, the marginal costs and
benefits of these regulations may be much closer than previously supposed.
This study did not attempt to derive an optimal DBCP standard using cost-benefit or other
criteria. As an illustration, however, assume that the marginal benefit of a risk reduction of 1 in 10,000 is
$200, a relatively conservative estimate.2 The marginal benefit and marginal cost of risk reduction are equal
at standards of about 8 in 1,000,000 for confidence levels of 95 and 99% and about 9.2 in 1,000,000 for
average risk.
dAnother way of expressing this is to say that the marginal cost of the good "safety" is sharply increasing.
GW2-31
-------
Session GW-2
Lichtenberg
6.0 FINAL REMARKS
The preceding discussion has argued the fact that the current knowledge and data collection
capabilities make health risk assessments subject to considerable uncertainty, while the sensitivity of the
public to uncertainty about health risks makes that uncertainty a central feature of regulatory decision
problems about ground water quality. The current method of dealing with uncertainty is to leave uncertainty
adjustment in the hands of risk analysts and to treat quantitative risk estimates as certain in cost-benefit
analyses. This approach is unsatisfactory because it involves inconsistent adjustments for uncertainty and
ignores the potential for uncertainty reductions in an optimal policy mix.
The safety-fixed approach proposed by Lichtenberg and Zilberman offers a practical
alternative with some attractive features, but deals only with cost. Applying cost-benefit criteria to obtain an
optimal risk standard requires: (a) fixing the confidence level arbitrarily and (b) assuming that the marginal
benefit of risk reduction is independent of the confidence level, (i.e., that individuals' willingness to pay for
risk reduction does not vary with the level of uncertainty about risk). Neither assumption is entirely
satisfactory. It would be preferable to set the confidence level according to decision-theoretic criteria, and
one would expect individuals to be willing to countenance tradeoffs between levels of risk and uncertainty
about risk. For example, if benefits were defined in terms of both the risk level R0 and confidence level a
(i.e., the benefits function was B(R0p)), then one would choose R0 and a to maximize net benefits B(R0p)-
C(RoA).
Such an approach to benefits differs fundamentally from the standard expected utility
framework, however, because it assumes that people have preferences over probabilities as well as over
outcomes. This implies that measures of the benefits of health risk reductions must be derived from a more
general approach to decision making under uncertainty than the expected utility model. Taking such a
direction seems justified by both observed behavior indicating strongly that uncertainty about health risk
matters, and by the observed anomalies on choice under uncertainty in a broader context.
7.0 REFERENCES
1. Jones-Lee, M.W. (ed.). The Value of Life and Safety. Amsterdam. North-Holland. 1982.
2. Fisher, A., L.G. Chestnut, and D.M. Violette. "The Value of Reducing Risks of Death: A
Note on New Evidence," Journal of Policy Analysis and Management. Vol. 8, pp. 88-100.
1989.
3. Lichtenberg, E. and D. Zilberman. "Efficient Regulation of Environmental Health Risks,"
Quarterly Journal of Economics. Vol. 49, pp. 167-178. 1988.
4. Lichtenberg, E. "Conservatism in Risk Assessment and Food Safety Policy," Economics of
Food Safety. New York. Elsevier. 1991.
5. Lichtenberg, E. D. Zilberman, and K.T. Bogen. "Regulating Environmental Health Risks
Under Uncertainty: Ground Water Contamination in California," Journal of Environmental
Economics and Management. Vol. 17, pp. 22-34. 1989.
6. Slovic, P., B. Fischoff, and S. Lichtenstein. "Facts and Fears: Understanding Perceived
Risk," Societal Risk Assessment: How Safe is Safe Enough? New York. Plenum. 1980.
7. Food and Drug Administration Pesticide Program. Residue Monitoring 1991. Washington,
D.C.. p. 17. September/October 1992.
8. U.S. Environmental Protection Agency. National Pesticides in Ground Water Survey:
Phase I Report. Washington, D.C. January 1989.
GW2-32
-------
Session GW-2
Lichtenberg
9. Machina, M.J. "Choice Under Uncertainty: Problems Resolved and Unresolved," Journal
of Economic Perspectives. Vol. 1, pp. 121-154. 1987.
10. International Commission on Radiological Protection. Report of the Task Group on
Reference Man. No. 23, pp. 11, 338-341, 358-360. New York: Pergamon Press, 1975.
11. Bailey. MJ. Reducing Risks to Life. American Enterprise Institute, Washington, D.C.
1980.
12. Broder I.E. and J.F. Morrall II. "The Economic Basis of OSHA's and EPA's Generic
Carcinogen Regulation," What Role for Government? Durban, NC. Duke University Press.
1983.
GW2-33
-------
Session GW-2
Michaels
When the Home Is No Longer a Castle:
Inferring the Economic Value of Ground Water Contamination
from Residential Property Sales
R. Gregory Michaels, Ph.D.
Senior Economist
Abt Associates, Inc.
ABSTRACT
Numerous incidences of hazardous contamination at sites around the country have generated strong reactions
by citizens living in the vicinity of such sites. Public health risks are usually at the forefront of these concerns
but other negative effects also get attention. One often expressed concern by households near a
contaminated site is the effect that the site has on the values of their homes.
Because of the apparent connection made by the public between observed economic behavior—the purchase
of housing-and the onset or discovery of contamination at a site, residential property values have been seen
as a source of information on the economic damages from ground water contamination. To generalize, the
property value approach entails using statistical methods to disaggregate housing prices or price changes into
different components-that part attributable to the influence of ground water contamination and all other
parts attributable to other factors, such as housing attributes, community characteristics, and housing market
conditions. This paper examines the experience to date with property value studies which have addressed
ground water contamination, directly or indirectly, and presents insights on the potential applicability of this
methodology to future studies of ground water valuation.
1.0 INTRODUCTION
"A man's home is his castle." The phrasing may be old but the sense of this saying still
captures some of the expectations that many households have about their homes. To them, the home is their
dominion, a place to retreat to, free of the onslaughts of the outside, workaday world. For most people, only
the home offers a place of privacy which does not have to be shared with society at large. Everywhere else,
at work, while traveling, we share space with the rest of society. As with many resources held in common,
this shared space is not always allocated well, with the result that congestion ensues. Faced with this context,
individuals are likely to value the privacy that comes with a home all the more.
Contrast this picture with the one wrought by environmental contamination in a residential
community. Homeowners feel threatened-the private space that was their own has been violated. While
homeowners may have immediate recourse when it comes to trespassing by people, they have little when it
comes to pollution. Contamination respects no property lines. As if to add insult to injury, not only do
private property rights offer little protection against this type of intrusion, homeowners may experience a
decline in the value of their homes as a result of the environmental contamination. The castle has become a
prison.
No wonder there is such an outcry from households in the vicinity of industrial sites and
waste management facilities contaminated with toxic material. Not only do they sense their physical well-
being at risk, one of their major capital assets is put in jeopardy. Anecdotal evidence is abundant that
property values are one of the major concerns of residents in the proximity of sites where environmental
contamination has been discovered. Ground water contamination is often at the heart of such discoveries. It
has such potential for surprise since underground contamination can go undetected for so long.
GW2-34
-------
Session GW-2
Michaels
As deplorable as the surprises are that come with the discovery of ground water
contamination, such scenarios have been observed with interest by researchers concerned with identifying the
economic damages associated with such contamination. Because of the apparent connection made by the
public between observed economic behavior—the purchase of housing-and the discovery of contamination
nearby, residential property values have been seen as a source of information on these economic damages.
Over the last fifteen years, a modest amount of empirical experience has been generated
which sheds light on the relationship between residential property values and ground water contamination.
The objective of this paper is to present insights gleaned from this body of work. Though this body of work
is primarily oriented to examining the effects of Superfund sites or hazardous waste management facilities, it
still has a bearing on other circumstances where ground water can be contaminated, such as the recent local
example of ground water contaminated by leakage from a petroleum tank farm. It is hoped that this review
will assist future studies of property values and ground water contamination move effectively from the
abundant anecdotal suggestions of property value effects to quantified and more verifiable estimates of
economic damages.
2.0 HOUSING MARKETS AND ENVIRONMENTAL CONTAMINATION
Given the ready concern that homeowners are willing to express at the appearance of
something they deem likely to damage the values of their homes, it would seem to be a simple matter to find
corroborating evidence for then- concerns in actual housing markets. To date, that has not been the case.
Circumstances which appear to be very prime candidates to provide this sort of empirical evidence have not
been uniformly successful. Oddly enough, this empirical experience does not necessarily contradict
expressions of community concern when contamination is discovered. This section sets out to explain how
the effects of contamination are expressed in residential property markets and therefore to illustrate why the
estimates of economic damages through property value studies to date have not always met expectations.
The first important clarification is that the impacts of contamination on housing values are
much better articulated in the news media in the immediate period after the discovery of contamination than
they are in the housing market. News happens fast but property markets move at a glacial pace by
comparison. Within days of the public announcement of major ground water contamination, it is not unusual
to find homeowners mobilized through existing networks, such as homeowner associations, to signal their
concern and discontent with the situation. Despite homeowners' sense of having been wronged, their
concerns will not be expressed in the housing market until homes in the vicinity of the site have been sold.
In ordinary times, the rate of sales may not be high-a dozen sales per month in an area with a few hundred
homes. After the revelation of contamination, this rate may even slow down in the near term as
homeowners hold onto their properties to see how the market reacts. While news articles might point to the
sale of a house at a tremendous discount as evidence of the property effects from the contamination, this
statistic cannot be treated as significant. It does, however, foster an intriguing hypothesis that there could be
property value effects, which can be formally examined once enough post-contamination sales have
accumulated.
The preceding discussion provides a good motivation for the issue of how the preferences of
homebuyers are expressed in the housing market. This issue will be examined through the elaboration of
different scenarios one could observe in the housing market following the discovery of ground water
contamination near one residential community in a large metropolitan housing market. Differences in these
scenarios are based upon different assumptions regarding the preferences of buyers in the housing market,
ranging from all buyers being very CONCERNED about ground water contamination (C) to all being GW
contamination NEUTRAL (N). This type of dichotomy is quite possible as McClelland et. al. showed in the
case of a Superfund landfill site in California.1 All individuals are alike in their housing preferences
otherwise. Total population is defined as P. References to a "seller" indicates the owner of a home in the
community affected by the ground water contamination.
GW2-35
-------
Session GW-2
Michaels
In the first scenario, the entire population (P) is composed of individuals who are concerned
about the contamination (C). Consequently, the housing market will reflect C preferences completely. There
will be a reduction in the price of homes in the affected community. Sellers in this community experience a
capital loss and therefore also a welfare loss. The buyers of homes in the affected community receive a
discount on the price of the home, relative to what they would have to pay elsewhere in the metropolitan
housing market. This discount reflects the economic value of contamination. This scenario is consistent with
the expectation that the housing market will pointedly reflect the economic impact from contamination.
In the second scenario, ah1 but one person in the population (P-l) are C and one person is
N. That one person happens to be a home seller in the affected community. Again, the market will reflect
C preferences. There will still be a capital loss by everyone selling a home hi the affected community,
including the individual with N preferences even though he/she is not concerned about the contamination.
Perhaps this explains why community reaction to the discovery of contamination can be so broadly supported,
since homeowners with N and C preferences alike have something to lose if they believe the rest of the
housing market has C preferences. Since that is assumed in this scenario, any buyer has to receive a
discount on the price of a home to reflect the economic value of contamination.
In the third scenario, the tables are turned. Everyone but one person in the housing market
(P-l) has N preferences. The one individual with C preferences is the seller of a home in the affected
community. Since the market will reflect N preferences, there should be no capital loss for this seller as long
as he does not advertise his contamination concerns, a step which would reveal his willingness to accept a
price lower than the going rate for all comparable houses. Although the seller is very concerned about
contamination, he experiences no welfare loss from a capital loss on his home. However, he still could
experience a welfare loss from any physical risks imposed by the contamination. Under this scenario, any
estimate of economic damages would have to be based upon risk assessment directly. The housing market
would offer no measure of economic damages.3
The fourth scenario is the most complicated and, not surprisingly, the most realistic. In this
case, it is assumed that half of the population has C preferences and half have N preferences. They are
evenly distributed throughout the existing housing market. Only a small part of the existing housing is
affected by contamination (e.g., less than 10%).
Once contamination is discovered, this mixture of preferences introduces a confusion into
the market which has to be sorted out. Half of the market values the contamination and the other half does
not. In the short run, there may be impacts on the residential property market while buyers and sellers try to
sort things out, such as a slowdown in the rate of home sales. Those sales which take place could include a
number of sales by C homeowners to N buyers, at a price greater than the C sellers would have been willing
to accept (equal to the price they would have been willing to pay for a comparably affected home) but less
than what the N buyers would have been willing to pay. In this case, there would be a partial capital loss
(relative to what would have occurred under the first scenario, where everyone has C preferences).
However, sales would also include homes owned by N individuals to N individuals, resulting in no capital
loss. As a result, the housing market would send mixed signals about the economic value of the
contamination. Identifying any property value effects becomes a challenge, unless a researcher can readily
identify sales involving C homeowners.
aThis scenario can be generalized to the case where everyone in the population has N preferences. Since no
one cares about the contamination, there will be no change in property values. Neither the welfare of the
sellers nor that of the buyers is affected by the contamination, as least as far as the value of homes is
concerned. Damage estimation again is left to the domain of risk assessment.
bHigher education and income levels may be one key to identifying households with such preferences.
GW2-36
-------
Session GW-2
Michaels
Under this same scenario in the long run, as long as the affected housing is a small part of
the market and there is competition among the N buyers, one would expect to see house sales reflect fully
what the N buyers are willing to pay. Homeowners hi the affected area who have C preferences will move
out. New buyers with N preferences will move in. In sum, there will be no capital loss expressed hi the
housing market but there will be a relocation of homeowners to accommodate then- preferences regarding
contamination.
From these scenarios, the following conclusion can be drawn. How much an individual
cares about contamination and how prevalent people with his preferences are in the housing market matters
a lot.c The first scenario offers the clearest circumstances where property value effects would be observed
in the housing market. However, circumstances where many individuals are substantially concerned about
the contamination may be rare. In part this can be explained by the fact that individuals respond differently
to the same information about environmental events. Some people care a lot, some care very little, at least
as expressed in subjective measures of risk. Differences in education, income, gender, number of children,
and age have been observed to be correlated with such differences in subjective risk.dl>2
Yet, the larger the contamination, the more likely it is that a large percentage of the
housing market would exhibit some concerns. The more obvious it is that there are major pollution
problems associated with a site, the likelier it is that a large part of the housing market is aware of it. When
substantial pollution emanates from a given site (e.g., discharges to several environmental media on a long-
term basis with well-publicized environmental impacts), the housing market can exhibit significant differences
between the prices of houses in an affected community and comparable ones in unaffected communities.
Whether ground water contamination itself constitutes a problem that would be considered substantial or not
is the subject of the next section.
3.0 MODELLING GROUND WATER CONTAMINATION IN PROPERTY VALUE STUDIES
Ground water contamination affects the welfare of individuals in several ways. Some of
these effects, such as potential health risks, can be linked to housing choices, some cannot, such as existence
values. Even where they can be linked, describing the connections has been challenging because not only are
physical processes involved, such as the movement of contamination underground, but also human behavior,
and most importantly, human perceptions.
It is tempting to think, for example, that the toxic risks from contamination should be
modelled in a manner analogous to the way it would be modelled in a risk assessment-characterize the
source and nature of the contaminant, its movement through the environment, the nature of human
exposure, and, ultimately, the resulting risks. This approach is not satisfactory because it presumes a level of
sophistication hi analysis and information which most individual homeowners are unlikely to have. It is more
likely that individuals will be much more simplified in then- assessment of the effects from contamination.
Accordingly, relatively sophisticated characterizations of the influence of contamination have not
outperformed simpler approaches in empirical property value studies. As a result, a large part of the
Environmental researchers do not escape this problem by focusing on measures which abstract from
preferences (such as risk assessment). It can be done only by making the strong assumption that the value of
a unit of risk is uniform across all people.
dldeally, then, a property value study would monitor the characteristics of the households who move out and
those who move in, to determine how differences in preferences toward contamination might contribute to
the ability or inability to observe price effects.
GW2-37
-------
Session GW-2
Michaels
modelling of the influence of contamination in these studies can be summarized in terms of three
attributes-proximity, timing, and area-wide or systemic effects.6
Where a single, identifiable source of contamination is involved, proximity of each house to
this source has become the common measure of the impact of contamination on housing prices. The
expected relationship is that housing prices increase with distance. Proximity has been depicted in various
forms in statistical specifications of the relationship between housing prices- using linear measures of
distance, nonlinear measures (including an exponential decline function intended to mimic the fate of
contaminants in the environment), and linear spline functions (which allow price effects to vary over discrete
intervals from the site). A review of property value studies to date, which have mostly focused on landfills,
Superfund sites, and hazardous waste management facilities, indicated housing prices increased by $900 to
$18,500 per mile of distance from the contamination source, with a median value of $3,800 (1991 dollars)/
These estimates reflect not only the influence of ground water contamination but other forms of
contamination as well. When limited to studies where ground water contamination was identified as a major
factor, the range became $3,000 to $9,000 per mile.
Identifying impacts on property values is sensitive to timing, in at least two ways. First,
when a distinctive contamination episode takes place (such as the discovery of long-standing leaks at a facility
or an accident which leads to immediate contamination), new information is provided to the public about the
nature of that facility's operations. Subsequently, public perceptions of the potential environmental impacts
from a facility are likely to change and these can translate into property value effects. Prior to such an
episode, the facility might have been expected to have some negative impact on local property values but
after the episode, the negative impact could be magnified. In most of the past studies of hazardous waste
sites, significant negative effects were identified after, but not prior to, a key contamination event.
Second, a contamination episode may not be translated into negative impacts on property
values, even if widely publicized. In a study of Superfund sites in Houston, Kohlhase found no impacts after
contamination was discovered until the sites had been added to the National Priority List.3 In other words,
only once governmental authorities had heightened concerns about the sites did they have a significant effect
on property values. Still, Kohlhase's results apply to the early 1980's. Twelve years into the Superfund
program, one would expect the public to have heightened sensitivities to the operation of hazardous waste
facilities, even though federal regulations have increased dramatically during the same period. The same
hazardous waste operations that had no impact in the 1970's and early 1980's could have a substantial impact
today.
Sensitivity to timing is one issue being investigated in a study of hazardous waste facilities
using ten years of residential property sales data, from the early 1980's to the early 1990's.6 In the
preparatory stages of this study, facilities where distinct contamination events could be identified became the
focus for analysis. At two of the three facilities being investigated, there are indications that the effects of
proximity to these facilities have grown over time.
While a powerful proxy for some effects of contamination sources on property values,
proximity does not capture certain area-wide effects associated with contamination. For example, the
contamination of public drinking wells, as occurred in the 1970's at the Woburn Wells Superfund site in
eThe insights presented are based upon property value studies of pollution from hazardous waste sites, such
as Superfund sites and hazardous waste management facilities, which includes but is not strictly limited to
ground water contamination.
fThese conclusions were based upon studies whose specifications were most amenable to expressing then-
results in terms of changes in housing value per mile of distance from the contamination source.2>3'4>5'6
Michaels7 calculated results for other studies in terms of percentage changes in housing price per mile.8'9'10
These ranged from approximately 1 to 7% per mile.
GW2-38
-------
Session GW-2
Michaels
Massachusetts, is one example where proximity may not be as important as systemic effects. Since these wells
provided 25% of the drinking water for the town of Woburn, more than the households living near the
contaminated sites were affected. A property-value model which tried to take these sort of systemic effects
into account found no economic damages for Woburn per se, but did for two other sites in Ashland and
Acton, Massachusetts, where ground water contamination was an important factor in their designation as
Superfund sites. The estimated damages for Acton and Ashland were $4.3 million and $47.6 million,
respectively.
A more aggregated approach to such area-wide effects has been realized through the use of
census data on individuals aggregated to the county level.11'12'13 Because of this aggregation, these models
cannot consider proximity to sites as an explanatory factor of differences in housing prices. Instead, the
models examine the influence of differences in county amenities and disamenities- including the number of
Superfund sites, the number of landfills, and the amount of waste disposed in landfills-on housing prices and
wages. While these studies did not consider ground water contamination directly, they probably include
Superfund sites where such contamination is present and thus provide at least a general impression of the
magnitude of estimates one could expect from this approach. The most recent of these studies found that
each household experiences annual economic damages of almost $750 (1992 dollars) on average. Expressed
in present value terms (using a 10% discount rate), the estimated damage per household is approximately
$7,500. This estimate appears to be similar to the average damages per household found in the site-specific
studies (i.e, those which estimate proximity effects), which ranged from $730 to $35,000 (median:
$6,300).S'h
The modelling experience to date offers general guidance on the circumstances where
property value effects are more likely to be identified and the magnitude of the findings in existing studies.
This discussion has focused on three factors that are particular to the way environmental contamination is
described. There are others, such as the income level and density of the homeowners in the vicinity of the
contamination source, which should be considered in the design of any property value study of localized
environmental disamenities.'
For lack of relevant precedents in earlier studies, this discussion has not provided extensive
insights into the way that ground water contamination alone can be evaluated through a property value
approach. Although several studies have dealt with sites where ground water contamination was present and
even a major concern, no studies have controlled explicitly for this type of contamination. To fill this gap in
understanding, any future property value study concerned with valuing ground water contamination should
begin with actual cases where the ground water medium is the predominant contamination target.J
glt was only possible to use a subset of the site-specific studies to calculate this range.2'4>5'14
It should be recognized though that the two approaches (site-specific and county-level) differ in the
populations to which their estimates apply. The estimates of the site-specific studies typically apply to
populations within one to four miles of a site. The county-level studies apply to all county residents. In
theory, the two population definitions could be reconciled in those counties where site-specific results could
be applied to or inferred for all sites in a county.
'Some of the other factors influencing the ability to identify property value effects from contamination are
discussed in.15
JIn this regard, incidents like the recent discovery of an underground leak of 175,000 gallons of fuel from the
Star Enterprise tank farm near a middle-income neighborhood in Fairfax County, Virginia may provide
useful circumstances for isolating the unique effects of ground water contamination on property values.
From the outset, property value impacts were identified as a matter of grave concern to nearby residents,
whose damage claims against the company were eventually settled in part through the company's provision of
GW2-39
-------
Session GW-2
Michaels
4.0 CONCLUSIONS
Directly or indirectly, most of the insights discussed in this paper hinge on how well
researchers depict how public perceptions of ground water contamination influence property values.
Precisely because those perceptions vary across homeowners and homebuyers and can vary across time for
an individual, the process of capturing the influence of these perceptions in an empirical model is so
challenging.
The challenges for any future studies do not stop with estimating damages, at least when
there is an interest in using the estimates to evaluate alternative policies to remediate ground water
contamination. These additional challenges also stem from the central role of perceptions. While the
recognition of damages from ground water contamination is instrumental to observing a negative impact on
property values, remediation has to reverse this process. Homeowners and homebuyers have to perceive that
remediation will eliminate the damages that they now perceive. Therefore, the test for the effectiveness of
remediation is greater in the realm of human perceptions than it is in the ground water medium itself. The
ground water could be rendered perfectly clean but this is for nought in the property market unless
individual perceptions change accordingly. Unless those who value ground water realize or believe that there
has been improvement, there will be no welfare gain. Again, the public's perceptions are everything.k
This paper has attempted to indicate how any future attempts to make inferences about
ground water contamination from residential housing markets can gain from the experience to date. That
experience is still too thin and the circumstances too varied to make conclusive assertions about the
advisability of inferring the economic value of ground water from property value studies. Property value
effects may not arise, even when the extent of physical contamination is extensive, and, even where they do
arise, the statistical obstacles are not slight. Nonetheless, anecdotal evidence and that from the existing
studies suggests that these effects are too large to be ignored.
5.0 REFERENCES
1. McClelland, G.H., W.D. Schultze, and B. Hurd. "The Effect of Risk Beliefs on Property
Values: A Case Study of a Hazardous Waste Site," Manuscript. University of Colorado.
1989.
2. Smith, V.K and W.H. Desvousges. "The Value of Avoiding a LULU: Hazardous Waste
Disposal Sites." Review of Economics and Statistics. Vol. LXVIII, No. 2. 1986.
3. Kohlhase, I.E. "The Impact of Toxic Waste Sites on Housing Values," Manuscript,
Department of Economics, University of Houston. 1988.
a guarantee of their housing values. This guarantee starts at 75% of the value the first year and graduates to
90% of the value by the fourth year. Although the company indicates this guarantee could cost $150 million,
if it accomplishes its goal of inspiring homeowners that something effective is being done, it could cost a lot
less. The housing market might even be on its way to recovery even before the contamination is fully
remediated. Ironically, this particular case is more useful without the settlement, which is substantial, than it
is with it since the conditional guarantee should mitigate most evidence of property value impacts hi future
housing sales (though not all, since the settlement was not for 100%). Accordingly, targets for any future
studies should include ones where settlement is either unlikely or not imminent.
kThese contingencies do not necessarily apply to all benefits from remediation. For example, when the
remediation of ground water reduces human health risks, ultimately individuals can benefit (because of
reduced medical costs, increased life expectancy, etc.) whether they are aware of the actual remediation or
not.
GW2-40
-------
Session GW-2
Michaels
4. Michaels, R.G. and V.K. Smith. "Market Segmentation and Valuing Amenities with
Hedonic Models: The Case of Hazardous Waste Sites," Journal of Urban Economics.
Vol. 28, pp. 223-242. 1990.
5. Thayer, M., H. Albers, and M. Rahmatian. "The Benefits of Reducing Exposure to Waste
Disposal Sites: A Hedonic Housing Value Approach," Manuscript. 1991.
6. Michaels, R.G., R. Byrnes, P. Giordano, W. Lin, and M. Wojcik. "Hazardous Waste
Facilities and Residential Property Values," Draft Report by Abt Associates Inc. to the U. S.
Environmental Protection Agency. 1992.
7. Michaels, R.G. "A Discrete and Continuous Choice Valuation of a Hazardous Waste Site's
Removal," Ph.D. Dissertation, Vanderbilt University. 1987.
8. Schultze, W., G. McClelland, B. Hurd, and J. Smith. Improving Accuracy and Reducing
Costs of Environmental Benefit Assessments: Volume IV. Estimating Benefits for Toxic
Waste Management: An Application of the Property Value Method. 1986. Draft Report,
U.S. Environmental Protection Agency. 1986.
9. Adler, K, R.C. Anderson, Z. Cook, R. Dower, and A.R. Ferguson. "The Benefits of
Regulating Hazardous Waste Disposal: Land Values as an Estimator," Report by the Public
Interest Economics Center to the U.S. Environmental Protection Agency. 1982.
10. Harrison, Jr., D. "Housing Values and the Willingness to Pay for Hazardous Waste
Regulations," Manuscript, John F. Kennedy School of Government, Harvard University.
1983.
11. Hoehn, J.P., M.C. Berger, and G.C. Blomquist. "A Hedonic Model of Interregional Wages,
Rents, and Amenity Values," Journal of Regional Science. Vol. 27, No. 4, pp. 605-620.
1987.
12. Blomquist, G.C., M.C. Berger, and J.P. Hoehn. "New Estimates of Quality of Life in Urban
Areas," American Economic Review. Vol. 78, pp. 89-107. 1988.
13. Walker, D.R., and J.P. Hoehn. "A Method for Estimating the Local Area Damages of
Superfund Waste Sites," Manuscript, Department of Agricultural Economics, Michigan State
University. 1992.
14. Harrison, Jr., D. and J. Stock. "Using the Hedonic Housing Method to Estimate the
Benefits of Hazardous Waste Cleanup," Research and Demonstration of Improved Methods
for Carrying Out Benefit-Cost Analyses of Individual Regulations. Vol. 1, Part I. Report
for the U.S. Environmental Protection Agency. 1984.
15. Michaels, R.G. "Prospects and Pitfalls: An Evaluation of Hedonic Housing Analysis
Applied to Superfund Sites," Paper presented at the Annual Meeting of the American Real
Estate and Urban Economics Association. New York, NY. 1988.
GW2-41
-------
Session GW-2
Questions and Answers
Questions and Answers
QUESTION: Has anyone on the panel done any analyses relating to the influence of
discernability, odors, taste and so forth on the rates of action or valuation?
DR. SMITH: Yes. There have been studies done both in the context of air and in the
context of water associated with discernability. Greg's comments earlier about the availability of this
information in the media and its effect on property values would be one example but there is an extensive
work linking to risk communication on this issue, too extensive to mention. I would be happy to provide it
afterwards.
QUESTION: Do you think CVM results are consistent with the observed consumer
behavior in the market?
DR. SCHULTZE: The main difference, in my judgment, is that respondents may not use
all of their energies to figure out true willingness to pay as they would do in actual transactions.
QUESTION: Do you differentiate rational respondents from irrational individuals? If you
do, why? So this is a two-part question.
DR. SCHULTZE: The question is, what is out there and is it comparable to a market
commodity? Let's differentiate between market commodities like Big Macs and non-market commodities
like cleaning up ground water. I would hypothesize that all of you know most of the characteristics of Big
Macs, including what they will do to your arteries, what they will do to your waistline, and just how
wonderful they taste, based on all of our youths. Remember your youth when you thought you could eat a
Big Mac safely.
So we know all of those things and it is pretty easy to come up with a value. I would submit
that the values in the lower of those two frequency distributions of values are the kinds of values we have for
Big Macs, a fairly tight distribution of values. All the evidence we have got is that values are log normally
distributed. That is a geometric or a logarithmic horizontal axis. When you know less about something, then
the variance, at least in a group, the variance in values will increase. Unfortunately it appears that there is a
lot of skew in that area of distribution that you do not see there because we have normalized it by taking a
logarithmic horizontal axis.
The mean of that upper distribution is substantially larger than the mean of the lower
distribution, so if we know more about something because there are skewed errors in value judgment, the
mean value will fall. I mean, how else would people buy cars like Corvettes when they are 18 years old?
They do not know about what it is like to drive a Corvette in a snowstorm. They have not thought about
that. Once you have tried to drive a Corvette in a snowstorm, or you realize there is no trunk and no place
to put children and other things, you change your mind and buy a Honda Accord.
Once you have learned something about automobiles or you pay the fuel bill or insurance or
theft. All of these other features that you did not think about when you bought that commodity. So there
are errors in market decision making.
So what I would submit, and here is the problem for nonuse values. If we go out and
survey a familiar commodity, I think we get very market-like values. If we are surveying for a very unfamiliar
commodity, then it becomes the burden of the survey itself to provide all of that information, and what we
see is this kind of collapse in the variance of values and a lowering in the mean value, and I think I am being
cut off. I will stop there.
GW2-42
-------
Session GW-2
Questions and Answers
QUESTION: How do you explain the high value and what implications are there for the
application of those values.
DR. BERGSTROM: The focus of our study was on protecting drinking water supplies.
For that focus of the study, in the definition of the commodity, we thought it was appropriate that what we
thought people would be interested in is whether the water is safe to drink or not, suggesting that the need
for a binary variable, either it is safe to drink, it is within EPA standards, or it is not safe to drink. We did
not really see where alternative levels there would be of interest or would be able to, we did not see how we
would convey that to the respondents and that is why we chose a binary variable, either all-or-nothing deal.
I am thinking in future studies it may be interesting to look at some kind of a ground water
quality ladder to look at different levels. For instance, you might start out with water that is safe to irrigate
crops in your gardens and moving to water that is safe to give to livestock and pets and water that is safe for
waste water purposes and then start getting to more contact uses like water that is safe for showering or
cleaning or cleaning pots and pans and finally the highest level being water that is safe for drinking or
cooking.
Again, what we were concerned is what are people willing to pay to preserve water quality
for drinking and cooking purposes and that binary variable is one reason why I think we got high values.
Another reason is in our without program situation, the questionnaire asks people to assume that there
would be no new treatment without this program so in our without program situation, again, it was kind of
an all-or-nothing deal that there would be, that if this program was not implemented, there would be no
other programs implemented either, so that was an important assumption.
Probably the most important assumption that we gave that I mentioned in my talk is we
asked respondents to assume that the program would be 100% effective. The risk literature suggests that
people highly value certainty. That when you are reducing risk and you get up to the point where you are
talking about 100 reduction in risk, the values really shoot up. The idea of certainty is very important to
people, and just put yourself in this position for a minute. If somebody came up to you and said, here is a
program that will guarantee that the water you drink will be safe, not just for the rest of your life but for
your children's life, your children's children's life, and your neighbors, it will protect the ground water for
your neighbors and your nieces and your nephews and all of that and it is going to end up costing you $50 a
month. That is what $600 a year would come out to.
It does not seem to me that that would be a totally unreasonable amount for people to pay.
The sample is skewed towards the higher income, higher education, older people. We need to do some
additional work to look at that, as Bill just mentioned, with that value, if we come back and do some better
adjustments for that, that may bring the mean values down. Embedding could be a problem where people
were perhaps getting us some values for environmental protection in general, that that could be a possibility,
although I am not sure that that is a real big problem. I do not think we understand a lot about embedding
yet, exactly what the implications are.
I think a real important question here, it will be the last one, it relates to what you were
saying about do not aggregate against over the entire population. So $640 per year per household implies
$160 billion nationwide. Is this credible? No, I do not think so. I do not think we would want to aggregate
it nationwide like that because as Gary mentioned earlier, aggregation had to ask the question, well, who
really cares about this issue and not everybody across the U.S. is going to care about this issue so the
aggregation can be much, much smaller.
DR. SMITH: An issue that is relevant to the points that John just raised, in thinking about
risk, people do like this notion of being safe, a threshold for safety and so it may well be very more difficult,
very much more difficult to explain degrees of safety, degrees of protection for different uses of water,
especially as they involve some amount of human exposure and, much more difficult than thinking about it,
saying the context of using water for recreation. In that context, economists have had some success in
GW2-43
-------
Session GW-2
Questions and Answers
explaining different degrees of water quality by the uses that were permitted. It is not as clear that that will
be true in the context of drinking, although, as John suggested, it remains a researchable question.
QUESTION: Were there any respondents who did nothing different to their water after
they were notified? If so, wouldn't this be your base level or lowest value?
DR. ABDALLA: Just to be clear, hi the central Pennsylvania study, 25% of the households
decided not to undertake any information even though they were aware. The second survey, southeast
Pennsylvania, it as about 55% who chose to do nothing even after they were informed of the risk.
So just to be clear there were a good number of people that differed by area. As to the
question about wouldn't this be your lowest base level, as I understand, the converting expenditures
approach. One thing I would like to point out, though, is the people who are not undertaking avoidance
actions still may be being exposed to health risks. Now, these may be a real long tune in terms of water
issues but with water-borne diseases, for example, people that are drinking the water are exposed to risks
and, in a short period of time they could feel those health impacts.
The second one is harder to answer. I will again try to be brief but it is very much related.
Why was the cost per week so different in the two study areas. In the central Pennsylvania site, it was 525
per week. In the southeast Pennsylvania site it was only $.40 per week. If I had more time in terms of
explanation, I would have pointed out that there is one, we handled treatment costs a little bit differently in
the studies, but, given that there were relatively few people that bought home water treatment devices, I do
not think that is enough to change the results that much.
My explanation or my answer is more directed at the context in each of these communities.
In one community there was a very high level of awareness of the problem. The media was very much
involved. I think 96% of those, the respondents to the survey, were aware of the problem. There was also
controversy. One reason there was controversy was there was no drinking water standard and you had
experts from various sources, it was a private water system supplying the water. I mean, investor-owned
company. They had one answer, the water was safe. There were some experts from the university saying it
was not safe so amid this confusion, I think that led to it, and I guess the issue of trust is another factor as
well. I think averting, what you do is probably related to how much you expect someone, the public, to do
for you and, in this case, this public water supplier did not have a good track record.
The area that came up with the much lower, the southeast Pennsylvania site, only $.40 per
week, there this contamination problem was being held in a very low-key way. Public notification was
minimal, sort of on the second page of a newsletter inserted with your water bill, and so you just had a very
different dynamic which I think affected averting expenditure levels.
GW2-44
-------
Session GW-3
Questions and Answers
Session GW-3: Policy and Regulatory
Considerations
SESSION SUMMARY
MODERATOR: Martha G. Prothro
PRESENTERS:
Velma Smiih-Policy and Regulatory Considerations
James A. Goodrich- Ground Water Value in California
Bill Weisrock— Ground Water Valuation: An Industry Perspective
Paul Jehn- Ground Water Valuation: The High Cost of Not Protecting Ground Water
Keith N. Color-Policy and Regulatory Considerations
Jimmie Powell-Policy and Regulatory Considerations
Debra Jacobson-/W/cy and Regulatory Considerations
GW3-1
-------
Session GW-3
Smith
Policy and Regulatory Considerations
(transcribed from audio tape)
Velma Smith
Director of Groundwater Protection Program
Friends of the Earth
I want to say "Make it simple" right at the start. Friends of the Earth believes, and I
believe, this is held by a lot of other folks in the environmental community, that ground water must be
valued and therefore protected as a resource. We think it is a grave mistake to value only the ground water
that we use today or even what we think we might use in the foreseeable future.
And before I explain why we feel so strongly on that point, I would like to take us from the
general to a few specifics. I would like you to consider in the short time we have this morning a couple of
particular ground water cases.
The first one is actually close by. It is only 12 to 15 miles west of here in the neighboring
county of Fairfax. Within the City of Fairfax, there is a 116-acre parcel of land that was zoned for residential
to industrial use back in the 1960's. That zoning change in an area that was growing, but clearly far less
populous than it is today, allowed development of what is known as the Pickett Road Tank Farm.
Four oil companies and a pipeline company operate from that site. Neighboring the tank
farm are commercial business areas and nearby residential developments. Now for a number of years in this
area, various folks in the residential developments had complained of gasoline odors. A lot of those
complaints were dismissed. But by this time, about two years ago, it became clear that there had, indeed,
been a leak of significant proportions from the tank farm.
Estimates of the amount of petroleum product that is moving through the fractured geology
of that area have ranged wildly, from 100,000 gallons to as high as 4.5 million. Three homes have been
evacuated, one for explosion dangers and two to allow for trenching, drilling, and other cleanup work to
proceed. Property values have plummeted and the affluent community in that neighborhood is in an uproar.
They are worried about their life's savings and their homes. They are worried about
uncertain risks to themselves and their children from breathing potentially toxic fumes. Upstanding,
conservative citizens have become frustrated, even to the point of being enraged. They are frustrated with
the mixed signals that they get about what is happening below the ground and who is responsible for fixing
the problem.
Normally quiet individuals can sit in a meeting room with public officials and shout their
anger, and citizen group leaders actually, a week ago, started talking about civil disobedience. But, wait.
Let's step back. If we are thinking about a national policy on ground water, let's consider what the ground is
used for in this area. Many years ago, then when undoubtedly people could tap the ground water for
drinking. The public water supplies from surface water have been available in this area for quite some time.
It would appear that there is simply no threat to drinking water supplies from the
underground plume of hydrocarbons out in Fairfax. Under federal policy that says we should focus our
protection and our cleanup efforts on usable ground water, we would have to draw the conclusion that we do
not really have a problem at Pickett Road. Let's move across the state. And I should say that I am not
picking on Virginia because I think that Virginia is worse than other states. I just happen to know a few
items from Virginia.
GW3-2
-------
Session GW-3
Smith
But if we move across the state to the coal region, in '85 the completion of a gas production
well had an unexpected result. Grout pumped into a newly drilled well traveled through the ground water to
a spring. That spring fed a state-owned fish hatchery. Thousands of trout were lost, thousands more had to
be transported to a clean water source.
Not a drinking water problem. And if one takes a look at EPA's ground water strategy and
its notion of ground water classification, we would say that it is clear that the connection to the trout
hatchery does not make what EPA calls a "highly valued" or "class one" ground water area. Those trout are
not endangered species, after all. Again, one might conclude from a national perspective that we did not
have much of a problem.
Outside of Virginia, but still on the east coast. The Green Point section of Brooklyn.
Gasoline explosion in the sewer system below Henry Street in 1950 told people that they had a problem. In
the 1970's, construction workers in the area halted excavation work when they encountered black material
oozing from the ground. Yes, oil again.
In 1978, New York officials traced the problem to oil storage tanks at a storage site along
Newton Creek. On three occasions between the '50's and the '70's, a local sewage treatment works had to
be shut down. The infiltrating oil killed the bacteria in the plant's treatment scheme.
Now, must we ask whether people in Brooklyn are drinking this ground water to determine
whether or not we have a problem? And if people were not drinking the ground water and were not
expected to drink the ground water, surely we would not have wanted the oil company to waste its scarce
capital resources on unnecessary pollution controls.
Friends of the Earth believes that a national policy, like some of those proposed by EPA,
that calls for valuing and therefore protecting ground water based solely on its potential as a drinking water
source, is wrong. It is short-sighted and can put at risk ground water sources that will one day be needed for
drinking. We do not have the ability to make good predictions about what we need, let alone what our
great-grandchildren will need.
Second, it is what I call "ecologically illiterate" and that is on two counts. First, the concept
of segmenting ground water resources into sections of which we will manage some for potable water and
some for free public sewers, simply defies natural science. Ground water is a dynamic moving resource. In
most cases, it moves quite slowly. But it does move and often in ways that are not entirely predictable.
Further, what we do on surface can change how ground water moves. We may be able to
spend enough money at a given site on consulting hydrogeologists to get a good snapshot of what is
happening below the earth's surface at a given point in space and time. But if ground water withdrawals
from increased consumption or from ground water pumping for a cleanup occur, then the dynamics of what
we have observed at that one point and time will change.
Managing rather than preventing contamination, in our view, is not only costly but probably
futile over the long term. Valuing ground water solely as a drinking water source is also ecologically
illiterate because it ignores the simple fact that ground water are surface water and inextricably linked.
Nearly one-third of America's streamflows originate as ground water and some estimates indicate that a
billion gallons of ground water seep into the Atlantic and Pacific Oceans every day.
So, what should we do? Can we really go about protecting this vast resource? I say, "Yes,
we can" and (just to borrow the phrase from Nike) we should "Just Do It." I think we can protect ground
water with appropriate changes to the Clean Water Act and to other federal legislation. We can move away
from antiquated and short-sighted views of ground water.
GW3-3
-------
Session GW-3
Smith
By taking advantages of lessons learned and programs built for surface water, we can
protect ground water and assure that we stop the shell game of moving pollution from one area to the other.
The important factors that we should call for is that we should call for a national goal to protect and
enhance the physical, chemical, and biological integrity of our nation's ground water resources across the
board, not just as drinking water supplies.
We need to recognize the connection with surface water. We cannot make the mistake of
waiting to control ground water discharges until we develop full and unassailable lexicological profiles for the
tens of thousands of chemicals in use. We cannot let standard setting set the pace for pollution. In our
view, the first line of defense in federal ground water policy should be source controls.
Effective ground water protection throughout the nation will occur only when the federal
government acts to adopt a national program on source controls to minimize or eliminate discharges of
pollutants and toxic chemicals to ground water. We believe there must be minimum standards of care and
control for sources and activities which have the potential to pollute ground water. Whether they are tank
farms, whether they are fertilizer applications, mine waste disposal, oil drilling, or septic tanks we must have
reasonable standards of care.
I am sure that many in the audience might disagree with these concepts. But what I hope
we can do on the ground water debate is move from broad discussions of general policy to actually debating
and hashing through specific proposals to protect ground water.
I might as well have my reference to last night's debate. And as I watched the debate, I was
thinking about the conference and about the debate on ground water policy in which I have a participant now
for almost 10 years. And I would like to say to Mr. Perot, "If you think that you have seen grid-lock hi
Washington, you haven't seen nothing yet until you look at what has been happening on ground water over
the last decade."
A heck of a lot of talk. A lot of hearings. A lot of conferences. A lot of policy dialogue,
but precious little action. Yes, we have some state ground water strategies and some states have jumped
ahead on certain problems and are to be commended for doing that. They have done good work in some
areas, but far from comprehensive and either from the perspective of geography or the range of
contamination sources out there.
And, yes, EPA may release yet another ground water strategy document, but I actually
would defy anyone here to tell me what that particular document will change in the real world. I hope I
have got you thinking.
GW3-4
-------
Session GW-3
Goodrich
Ground Water Value in California
James A. Goodrich
Executive Director
San Gabriel Basin Water Quality Authority
ABSTRACT
In California, water is free! Though a price can be affixed to the commodity due to transporting costs, its
true value lies in the ability of the storage facility to hold sufficient water to meet demands through extended
dry periods, such as those occurring today in California. Because surface water storage capacity is
insufficient to meet demands through dry periods, many areas must rely on vast ground water storage basins.
The significance of this has become painfully apparent to many who must rely solely on surface water during
this sixth year of drought. The San Francisco Bay area, which receives most of its water from surface
storage systems in the Sierras, must meet demands through rationing. On the other hand, if the Los Angeles
metropolitan area did not have huge underlying ground water reserves to successfully fend off the drought, it
would have been forced to augment its dwindling surface water supplies with desalted sea water at a cost of
$2,000 per thousand cubic meters (per acre-foot). In California's Central Valley, the drought has forced
farmers to switch from low-cost surface water to overdrafting their ground water basins by 15 billion cubic
meters (12 million acre-feet) or more in order to stay in business. Therefore, like an insurance policy, one
only realizes the true value of ground water storage in times of dire need.
1.0 INTRODUCTION
The value of ground water in California's arid and semi-arid climate is not in the cost of the
commodity, but is hi the size of the available storage capacity of the ground water basins relative to the long-
term demands for water. The coordinated (conjunctive) use of surface water with ground water further
enhances the value of both resources. And, as both human and environmental demands grow, this
coordinated use of resources will become critical to the economic and environmental health of the state.
Therefore, the value of ground water cannot be determined without also determining its interrelationship to
surface supplies in the very complex water resources structure of the state.
This discussion will briefly review the gross hydrology of the state, the characteristics of the
water supply system in California, some of the costs associated with delivering water to its points of demand,
problems associated with a limited supply with competing uses, and finally, how conjunctive use can enhance
the reliability of long-term supplies.
Precipitation yields on average about 247 billion cubic meters (200 million acre-feet) of
water annually hi California. Of this amount, most occurs in the Sierra Nevada Mountains in the form of
snow pack, which ranges from 125 to 250 centimeters (50 to 100 inches) in depth. Approximately two-thirds
of this precipitation is lost to evapotranspiration (of non-human influence), while the remainder occurs as
runoff. The snow pack, in a sense, is used as a vast reservoir to meter water in the spring to manmade
surface reservoirs and ground water basins. Approximately 46 billion cubic meters (37 million acre-feet) are
captured and used by man, of which about 40 billion cubic meters (32 million acre-feet) is used by the
agriculture industry and 6 billion cubic meters (5 million acre-feet) goes toward municipal and industrial
uses.
The state has 152 major surface water reservoirs with a total storage capacity of 53 billion
cubic meters (43 million acre-feet). The majority of this water is delivered through a series of large
aqueducts and canals. Some of these reservoirs, aqueducts, and canals are illustrated on Figure 1. The 450
GW3-5
-------
Session GW-3
Goodrich
recognized ground water basins in the state have a total storage capacity of 1 trillion cubic meters (850
million acre-feet), of which about 494 billion cubic meters (400 million acre-feet) are recoverable. The
average annual safe yield of ground water basins without significant conjunctive use is about 18.5 billion cubic
meters (15 million acre-feet) per year.
2.0 THE COST OF WATER
In California, water is free. Costs are associated with storing and delivering the water to
points of use. For example, water from the federally owned Central Valley Project (CVP) delivered to
farmers in California's San Joaquin and Sacramento Valleys (together forming the Central Valley) is between
$20 and $30 per thousand cubic meters (acre-foot). The cost of this water is partially subsidized by earnings
on the generation and sale of hydroelectric power. Further, the debt used to finance this project in the 1930s
has been retired. The State Water Project (SWP), owned and operated by the state, is a newer facility and
has a cost to Central Valley farmers of $100 to $200 per thousand cubic meters (acre-foot) due to debt
retirement obligations. The SWP water delivered to southern California must be pumped over a mountain
range, resulting in an additional $20 to $40 per thousand cubic meters (acre-foot) cost.
Ground water left in storage is also free. However, the cost of pumping it to the point of
use ranges between $20 and $100 per thousand cubic meters (acre-foot), depending on the depth of the
water table and the elevation of intended use. In a heavily used ground water basin, the price of ground
water increases due to several factors. First, extractions beyond natural safe yield must be replenished by
imported sources. The cost of imported surface water from the SWP or CVP is discussed in the previous
paragraph. Add to the imported commodity cost the cost of building recharge facilities over the ground
water basin and conveyance systems to get imported water to the recharge facilities. These recharge facilities
can be in the form of surface water spreading systems like those used in southern California or injection
wells. For example, recharging water in surface spreading facilities in Orange County costs between $10 and
$20 per thousand cubic meters (acre-foot), exclusive of the cost of the imported water.
In coastal areas, heavily used ground water basins must maintain barriers to sea water
intrusion. The annualized capital cost of injection systems, which can only use highly filtered water, is about
$100 per thousand cubic meters (acre-foot). Imported water or reclaimed wastewater can be used for
injection. The cost of reclaiming wastewater suitable (by public health standards) for injection ranges from
$300 to $600 per thousand cubic meters (acre-foot). However, the cost impact on the ground water resource,
as a whole, is substantially less because only a few million cubic meters (thousand acre-feet) of this high cost
water is required to protect a basin containing several tens of millions of cubic meters (several hundred
thousand acre-feet) or more of potable ground water.
If ground water contains naturally occurring or man-made contamination, the cost of using
this source is further increased. Most organic solvents, for example, cost about $75 to $100 per thousand
cubic meters (acre-foot) to remove. Nitrates cost over $200 per thousand cubic meters (acre-foot) to remove
to drinking water standards.
3.0 DROUGHT IMPACTS ON CALIFORNIA'S WATER RESOURCES
The value of ground water in California is in the size of the storage capacity of ground
water basins. As stated earlier, the average demand for water in the state approaches 50 billion cubic meters
(40 million acre-feet) per year. Surface storage is a little more than this. Therefore, any extended dry
period can quickly outstrip the ability of surface storage systems to meet their intended demands. Current
planning efforts anticipate periodic shortages lasting, at most, four to five years. However, for the
Sacramento-Delta area, the worst drought in the twentieth century had a duration of seven years (1927-1934).
We are now entering the seventh year of below normal precipitation in California. Tree ring studies indicate
droughts have lasted as long as 60 years (1760 to 1820) in California. Though a 50-year drought would
probably be catastrophic to both the state's economy and biological resources, planning and surviving a
GW3-6
-------
Session GW-3
Goodrich
drought of 10 to 15 years duration is not an unreasonable objective if water resources are properly allocated
and managed. However, maximizing the utility of ground water resources is critical to meeting this objective.
Today, entering the seventh year of the drought, the impacts on various communities has
varied. San Francisco receives most of its water supplies from the Hetch Hetchy Reservoir located just north
of Yosemite Park in the Sierra Nevada Mountains. The reservoir is nearly depleted, forcing the Bay area to
ration its remaining supply, and turn to limited local ground water supplies for some uses. Some
communities in the area are looking to sea water desalting to provide some base level of water to meet their
demands during periods of drought, at a cost of nearly $2,000 per thousand cubic meters (per acre-foot).
Farmers in the Central Valley area, facing a 90% cut in supply from surface supplies, have
turned to more expensive ground water to remain in business. This has resulted in more than a 15 billion
cubic-meter (12 million acre-foot) overdraft of Central Valley ground water basins. When the weather turns
wet again, it will take several years of non-pumping to recover from the overdraft conditions. Only with
diversion and recharge of large quantities of surplus surface water during wet periods can these basins
recover hi time for the next drought.
In southern California, surface supplies delivered from the State Water Project have been
cut back. To meet demands, water agencies have increased their take from the Colorado River and from the
large coastal ground water basins. The ability of southern California to tap the Colorado River to this extent
will not be possible in the future as Arizona brings Colorado River water through its newly constructed
Central Arizona Project. This leaves the water available in ground water basins as the only buffer for future
droughts.
Man is not alone in suffering the impact of the current drought. Reduced flows in the
Sacramento-San Joaquin River Delta, located between San Francisco Bay and Sacramento, have severely
impacted fisheries, particularly striped bass and salmon. Many biologists predict that other organisms face
crisis in this ecosystem should the drought continue.
4.0 CONJUNCTIVE USE OF SURFACE AND GROUND WATER
To avoid the economic and biological catastrophe of extended droughts in California will
require better coordinated use of surface water and ground water supplies. Meeting this objective of making
the state "drought resistant" will require widespread conjunctive use of surface and ground water. Two
operational forms of conjunctive use must be differentiated. The "put and take" or "seasonal storage" form of
conjunctive use requires that an amount of water be recharged and extracted on an annual basis, thereby
augmenting the annual yield of the basin. This form of conjunctive use minimizes over burdening either the
surface water or ground water delivery systems during the course of a season. The "long-term" form of
conjunctive use consists of banking large quantities of water during times of surplus surface water for the
purpose of alleviating overdraft ground water conditions created during periodic water shortages, or
droughts. The goal would be tq divert surplus surface water during wet and normal runoff years and store it
underground closer to the points of use. To fully utilize the stored ground water, well facilities would have
to be modified to allow extraction of water from deeper aquifers hi extremely overdrafted basins. Surface
water delivery systems must also be expanded to allow normal demands to be met, as well as delivering
surplus water to points of ground water recharge. In addition, facilities must be built and operations
modified to rapidly recharge overdrafted basins when surplus surface water is available.
5.0 THE VALUE OF GROUND WATER
As stated earlier, water stored in surface water systems can be easily depleted hi one to
three years of below normal runoff. Because of the vast amounts of ground water stored in California
basins, demands during drought conditions can be met generally for a much longer period (up to 15 years)
than with surface water storage systems. But, once depleted, ground water systems take many years to
replenish, depending on the amount and tuning of local runoff (i.e., runoff spread evenly over several months
GW3-7
-------
Session GW-3
Goodrich
versus all runoff for the season coming in a one-week period). Combining the rapid storing and diversion
capability of surface water systems with the extremely large storage capacity of ground water basins can
minimize the loss of fresh water to the ocean and enhance the yield of both systems, particularly during
periods of drought. This is the objective of surface water and ground water conjunctive use operations in
California. In summary, the demands must be met by redundant surface water and ground water storage
and delivery systems to take full advantage of limited supplies during this period of growing demands and
ever-changing climatic conditions.
Therefore, the value of ground water should not be measured in the terms of the cost of the
commodity, but in terms of the storage capacity that can be used during times of drought. The yardstick by
which one should measure the value of ground water in an area should be in the potential economical and
biological losses incurred during a drought if that ground water basin is not available due to overdraft or
degradation by contamination. So, like an insurance policy, one only realizes the true value of ground water
storage in times of dire need.
GW3-8
-------
Session GW-3
Goodrich
ERA CANAL
Miller ton Lake
Figure 1.
alton S«»
\ [imptrlml
""
Major State and Federal
Water Development in
California
All Ammrtetn
SAN OIEOO Cenel
GW3-9
-------
Session GW-3
Weisrock
Ground Water Valuation; An Industry Perspective
William P. Weisrock
Director, Ground Water Management
Amoco Corporation
ABSTRACT
Industry recognizes the value of our nation's ground water resources and the need for their protection.
Many corporations have adopted proactive ground water protection programs, and support advocacy efforts
through trade associations such as CMA and API. Industry expenditures for pollution prevention and
remediation of existing ground water contamination are steadily increasing. However, many in industry share
a belief that ground water resources can and should be managed in a more cost-efficient manner.
Only about 10% of total U.S. ground water withdrawals are used for industrial purposes, primarily for non-
contact cooling. Given this relatively low usage, the overall cost to industry to obtain and to treat ground
water is relatively small. More importantly, many industrial facilities are located over or adjacent to aquifers,
and the costs associated with protecting these ground water supplies and restoring them through remediation
can be substantial. Therefore, from a valuation standpoint, the costs of protection and restoration of ground
water are of primary concern to industry.
This paper raises a number of issues that should be, but often are not, considered when attempting to place
a value on ground water in terms of protection and remediation.
On the protection side, the paper discusses cost-benefit approaches, the need for sensible ground water
classification standards, detection and monitoring costs, and the impact of local or site-specific conditions on
costs and valuation.
On the remediation side, ground water valuation is strongly affected by remediation costs, time requirements
for aquifer restoration, and limitations of remedial technology. This, coupled with universally inherent
financial resource limitations, suggests that risk-based approaches to remediation and cleanup level targets
are needed. Risk based approaches should consider protection of human health, the beneficial use of the
impacted ground water supply, the cost of restoration to that beneficial use, and the cost to obtain or supply
alternative sources of water.
1.0 INTRODUCTION
This paper offers an industry viewpoint on the value and valuation of ground water. As a
starting point, it is fair to say that industry generally recognizes the value of our nation's ground water
resources and the need for then- protection for beneficial uses. Many corporations, including Amoco, have
adopted proactive ground water protection programs. Numerous trade associations such as the Chemical
Manufacturers Association and the American Petroleum Institute have instituted advocacy efforts on ground
water protection. Moreover, industry expenditures for pollution prevention and remediation of existing
contamination are steadily increasing.
However, many in industry share a belief that ground water resources can and should be
managed and protected in a more intelligent and cost-efficient manner. This paper suggests some
approaches to accomplish this.
As an aside, although industry is usually viewed in a corporate or monolithic sense in
discussions of environmental concerns, recall Pogo's First Law: "We Have Met the Enemy and He Is Us."
Industry is composed of people, many of whom use ground water hi their daily lives. As more and more
GW3-10
-------
Session GW-3
Weisrock
people in industry are made aware of the value of ground water and of their impact on ground supplies, and
take responsibility to change industry practices to protect those supplies, overall ground water quality should
begin to improve. Education of employees and the general public is key in this regard.
By the same token, we all share the benefits of the goods and services that industry
provides. We also share in making difficult choices and trade offs to determine how these benefits can be
sustained at a reasonable cost while properly preserving and valuing existing and future ground water
supplies.
2.0 INDUSTRY VALUATION OF GROUND WATER
Consider an industrial facility located over or adjacent to an aquifer. Depending on the
nature of the operations, this facility may or may not use the ground water in the aquifer. If it does, it will
assign one or more costs and one or more values to the ground water. The ground water may be valued as
an ingredient in a process, such as brewing, as a process solvent or diluent, or as a medium for disposal, as
for example, through a permitted Class II or Class V injection well.
Conversely, there are four other valuation or cost factors that the facility could consider:
Cost of = Cost to + Cost to + Cost to + Cost to
Ground Water Obtain/Produce Treat/Use Protect Restore/Replace
The first two cost factors, it can be argued, tend to be fairly small, particularly since
treatment requirements for many industry uses, such as once-through cooling water, are usually minimal. In
addition, as shown in Figure 1, only about 25% of industry needs are supplied by ground water, with the
dominant use being for cooling. According to O'Neil and Raucher,1 only about 10% of ground water
withdrawals are related to industrial demand.
On the other hand, what about the costs to a facility to protect the ground water residing
beneath or adjacent to it and the cost to restore it if contaminated by industrial operations? These costs can
be very significant depending on the use, value and vulnerability of the aquifer, the nature of industrial
operations, and related regulations. These costs are also interrelated: as the costs of protection increase, the
future costs to restore or remediate should go down, due to a decreased probability of contamination.
Given the potential magnitude of the costs to protect and restore ground water and the
current focus on ground water protection within EPA and the states, this paper will focus on these two
aspects of ground water valuation.
2.1 Ground Water Protection Aspects
First, let's consider protection of ground water supplies and the attendant cost. There is no
question that industry will continue to expend considerable resources, both financial and human, to develop
and maintain both proactive and compliance-driven efforts to prevent ground water contamination and to
reduce future liabilities. The maxim "pay now or pay later" is very true in this regard.
However, in trying to manage "up front" protection costs and develop a reasonable approach
to protection of current and future supplies, industry and the regulators should consider several cost factors.
2.1.1 Cost/Benefit Analysis
First, not enough emphasis is placed on cost/benefit analysis. Calculating the economic
advantages of near-term expenditures for ground water protection in terms of reduced future remediation
GW3-11
-------
Session GW-3
Weisrock
costs and liabilities is difficult, and is often not done. Intuitively the benefits are there, but they are not
readily quantified. One can consider three possible approaches:
1. When scenarios are developed to project the add-on costs of environmental
regulations, one should consider whether an additional expense for prevention
measures might result in a net reduction in those regulatory costs.
2. Case studies involving actual incidences of ground water contamination could be
analyzed to compare the costs of assessment and remediation with back-calculated
costs for protection which would have been needed in order to avoid the problem in
the first place.
3. Data should be developed to clarify which types of ground water protection actions
effectively reduce the need for remediation, particularly at older facilities. This
would also require a case-study approach.
Use of cost-benefit approaches will not be foolproof, because with any protective measures
there is still a finite probability that point source contamination will occur. However, generation of such
information would provide a better framework to understand the costs and potential cost savings associated
with ground water protection. It would also aid in determining the relative cost to protect aquifers with
different beneficial uses.
2.12 Beneficial Uses of Aquifers
The concept of beneficial use leads to a second-cost consideration in ground water
protection. Many aquifers are unsuited for drinking water purposes due to natural causes, such as salinity,
hardness, iron, etc., and it is estimated that about 60% of all ground water in its natural state is unsuitable
for drinking,2 due to problems with quality or deliverability. Much of this same ground water may be
entirely appropriate for agricultural or industrial purposes. It is not necessary to require that such ground
water meet or be protected to drinking water standards. Therefore, industry supports classification schemes
for ground water which accurately reflect the current and future beneficial uses and the vulnerability of the
ground water supplies. Industry also supports measures whose level of sophistication and costs are consistent
with the ground water resource value and sensitivity to contamination. Advocacy of a classification scheme
and risk-based protection does not constitute "institutionalized degradation—rather it supports the concept of
effective resource management.
2.13 Detection and Monitoring of G/ound Water Quality
Another major cost factor in ground water protection is detection of contamination and
monitoring of ground water quality. There is no question that monitoring and detection are needed.
However, blanket application of sampling and analysis requirements and regulations may be unnecessary
depending en the aquifer characteristics and beneficial use, and. sampling overkill may increase protection
costs significantly. Proper consideration of site-specific conditions can keep these costs reasonable while
adequately protecting ground water resources.
As an example of the complexity of this issue, Amoco recently conducted a joint Pollution
Prevention project (Figure 2) with EPA ?.t our Yorktown, Virginia, refinery.3 The data-gathering portion of
this study included a comprehensive study of ground water quality beneath the 243 ha. (600 acre) refinery.
The cost of this portion of the study alone, which included drilling 50 new monitoring wells in addition to the
94 existing wells, and analyzing for over 140 potential ground water pollutants, exceeded $150,000 and was
time consuming. This was to get data for only one point in time for a relatively young, medium-sized
refinery (about 35 years old, 8,500 m3/day {53,000 bbl/day} capacity).
GW3-12
-------
Session GW-3
Weisrock
There are several key points from this study regarding ground water monitoring:
1. The impact of the refinery operations on ground water was limited, even after 35
years of operation.
2. Even with a significant commitment of time and dollars, only selected sources in the
refinery could be sampled and at only one point in time.
3. Many chemicals of possible concern were at nondetect levels, and further
monitoring for these would have been pointless.
4. Shallow ground water beneath this facility has no current or potential beneficial use.
The bottom line is that significant resources are often expended to characterize ground
water quality and aquifer characteristics. Rather than imposing blanket requirements for detection and
monitoring, consideration should be given to the use, value, and vulnerability of the resources, as well as the
site-specific nature of the aquifer and the type of operations at a facility.
2.1.4 Application of Ground Water Protection Measures
In addition to monitoring and detection, across-the-board requirements for ground water
protection measures may not always be appropriate, without regard to site-specific conditions. To use
another example from the Yorktown study (Figure 3): a universal past practice at refineries and chemical
plants has been to install underground sewers to capture process fluids for disposal. Due to their age and
construction, many of these process sewers leak. A common practice for ground water protection in recent
years has been to install above-grade process sewers. However, the situation at Yorktown is somewhat
different. The process sewers, because of the height of the water table and the local hydraulic gradients,
tend to capture contaminated ground water and carry it to the wastewater treatment system. In effect, the
process sewer is acting as a sink, not a source, for benzene-contaminated ground water and is really
functioning as an in-place remediation system. In this instance, installation of above grade sewers and
decommissioning of the underground sewers would remove this remediation system.
22 Remediation Aspects
The cost factors involved in remediation or restoration of ground water resources typically
include:
• Assessment costs (to determine the extent of contamination).
• Engineering costs (to develop and install a remedial system).
• Remediation costs (to remove and/or destroy the contaminants).
• Any fines or penalties assessed.
Proper valuation of a ground water resource in terms of remediation costs should consider several questions:
1. Is the value of the ground water or aquifer worth the cost and time needed to
restore it, given the limitations of the technology?
2. Is the value of the aquifer worth the delay usually involved in beginning to restore
it?
GW3-13
-------
Session GW-3
Weisrock
3. Is the value of the aquifer more consistent with future restoration in combination of
point-of-use treatment?
22.1 Technology and Cost Limitations
Consider these data (Figure 4) on reduction of dissolved benzene in ground water at a site
undergoing pump and treat remediation. At approximately 200 to 300 ppb, the reduction in concentration of
benzene becomes asymptotic (levels off) for various reasons, including slow continued release of
hydrocarbons trapped in the aquifer matrix. This asymptotic concentration is 40 times higher than the 5 ppb
MCL for benzene in drinking water and 200 times greater than the 1 ppb clean-up standard being considered
by some states. Figure 5 graphically illustrates the converse of this, namely that clean-up costs increase
exponentially depending on the cleanup level targeted.
222 Regulatory Delays
Administrative and regulatory processes can significantly impede the initiation of
remediation efforts whether through protracted site assessment requirements, permitting delays, or review
and concurrence on remedial designs. Rather than treat them all equally, higher value aquifers should be
identified and given preferential treatment leading to faster remediation inception. This would not only
reduce costs, but would likely result in treating a smaller problem.
223 Future Vs. Current Remediation
Finally, under the Safe Drinking Water Act, all aquifers containing less than 10,000 ppm
total dissolved solids (TDS) are considered potentially potable. If contamination of such aquifers occurs, the
cost of near-term remediation may not make economic sense, given that additional costs will be incurred in
the future anyway to meet the U.S. primary drinking water standard of 500 mg/liter. In addition, future
R#D developments in remediation technology may lower the total overall cost and point-of-use treatment
may be more effective. This does, of course, assume that initial efforts are undertaken to control or
eliminate the source of aquifer contamination, to eliminate any free-phase hydrocarbons present, and to
contain the plume.
3.0 SUMMARY
This paper has attempted to develop a qualitative industry perspective on ground water
valuation which emphasizes the main industry concerns of protection and remediation of ground water
resources and their attendant costs. Industry's perspective is that environmental stewardship is not and
cannot be a "maximalist" philosophy whereby one does everything possible to protect a resource such as
ground water without regard to the costs or benefits. Rather, the very real factors of cost, timing, and
technology involved in ground water protection and restoration argue for a risk-based approach to ground
water valuation. A risk-based approach needs to consider (Figure 6):
1. The nature, toxicity, and extent of the contamination;
2. The potential for exposure and pathways by which exposure may occur;
3. The beneficial uses of the ground water;
4. Available treatment technology and engineering limitations;
5. The cost of remediation to restore ground water for that beneficial use; and
6. The cost to obtain or supply alternative sources of water.
GW3-14
-------
Session GW-3
Weisrock
Such a risk-based approach is entirely consistent with the maintenance of potable water
aquifers and the protection of human health, and indeed would assign the highest priority to these.
However, given universally inherent financial resource limitations, protection and remediation of all potential
sources of potable water as well as nonpotable ground water supplies becomes more problematic.
4.0 REFERENCES
1. O'Neill, W.B. and R.S Raucher. "The Costs of Ground Water Contamination," J. Soil and
Water Conservation, p. 180. March-April 1990.
2. Field, S.D. "Ground Water Protection and Reclamation," J. Env. Engineering. Vol. 116,
No. 4, p. 654. 1990.
3. Klee, H. and M. Podar. Amoco-USEPA Pollution Prevention Project. Yorktown. Virginia.
NTIS No. PB92228519. 1992.
4. American Petroleum Institute. "Technology Limits of Ground Water Remediation: A
Statistical Evaluation Method," API Publication No. 4510. 1991.
GW3-15
-------
Session GW-3
Weisrock
100
H 6(H
03
a! 4
-------
Session GW-3
Weisrock
£g -VI Separator;
I f. /
'&".
i
/ AMOCO Ol L COMPAN Y
'<^, A \ , I Iri
f/..- \ f * J « t;
/// ^' \ r °c ° I b
Ax \n i.
y>, j^—_
;' V
\ /If J
'j X"^"**8*8*-^*' POLLUTION PREVENTION PROJECT
,X X AMOCO CORPORATION
' GROUNDWATER MANAGEMENT SECTION
Figure 3
Yorktown Site Map Showing Process Sewer
1600
3* 1400,
§; 1200 n
s
CO §
o
§
o
1000 -
800 -
600 -
400 -
200 -
0
Figure 4
*****
100
200
300
400
Time in Days from First Measurement
500
Technological Limits to Remediation
GW3-17
-------
Session GW-3
Weisrock
Benzene
(Air Stripping + Free Hydrocarbon Recovery)
100 200 300 400 500 600 700 800 900 1000
Dissolved Hydrocarbon Concentration (ppb)
Figure 5
Total Aquifer Restoration Costs as a Function of Aquifer
Restoration Level
Risk - Based Approaches Should Consider:
• Nature, toxicity, extent of contamination
• Potential for exposure
• Beneficial uses of ground water
• Available treatment technology
• Cost of remediation
• Cost to obtain / supply alternative water sources
Figure 6
Considerations of Risk - Based Approaches
GW3-18
-------
Session GW-3
Jehn
Ground Water Valuation
The High Cost of Not Protecting Ground Water
Paul Jehn
Associate Director
Idaho Water Resources
Research Institute
University of Idaho, Boise Center
Nancy Bowser
Idaho Department of Health & Welfare
Division of Environmental Quality
Planning and Evaluation
ABSTRACT
The nation's ground water is a valuable natural resource and requires protection for current
and future beneficial uses. These beneficial uses include drinking water, industrial uses, agriculture, food
processing, and aquaculture. Protection of ground water for the many identified beneficial uses frequently
requires protection to at least drinking water standards.
In many of the western states, ground water is the major source of drinking water and forms
the base flow of rivers and lakes. Over the past 10 years, it has been shown the practices designed to protect
surface water are known to degrade ground water. In some cases this degraded ground water is recharging
and now degrading surface water.
The cost of supplying drinking water from ground water increases dramatically if ground
water quality is threatened or impacted either by insufficient prevention activities or upgradient
contamination. The increased monitoring requirements can be cost prohibitive to a small water system. The
remediation of contaminated ground water is expensive and frequently impossible.
The treatment of contaminated drinking water can cost water purveyors, and therefore the
consumer, hundreds of thousands of dollars. One new water supply well, if an uncontaminated source can be
located, can easily cost $500,000 or more.
Ground water is also used as media for the discharge of water from consumptive uses.
Virtually all activities (septic systems, agriculture, and urban development) have the potential to, and do,
impact ground water quality. Nondegradation of ground water is impossible. However, it is possible to
manage land use activities to ensure that ground water remains of drinking water quality. This does not
imply that ground water can be degraded to the drinking water standard.
1.0 INTRODUCTION
There is considerable debate occurring nationally regarding how much protection should be
given to ground water. The term "reasonably expected sources of drinking water" has caused a considerable
amount of confusion amongst congressional staff, policy makers, regulators and the public. This statement
implies, for some, that water can be degraded to drinking water standards and it has been suggested that if
water is not currently being used for drinking water, nor is expected to be used in the future, that ground
water can then be "written off. However, since drinking water is only one of many beneficial uses of ground
water it is imperative to continue protection of ground water for other beneficial uses.
GW3-19
-------
Session GW-3
Jehn
The amount of ground water protection required depends on the identified beneficial uses
which include: drinking water, surface water recharge, aquaculture, agriculture, industry, and as a media for
disposal of wastewater. The question that now arises is, how much it is worth to maintain the current and
future beneficial uses?
2.0
NONDRINKING WATER USES OF GROUND WATER
Nondrinking water beneficial uses of ground water can require water of equal or higher
standards than are required for drinking water. Examples of selected beneficial uses include aquaculture,
surface water, recharge, mining, mineral processing and as a disposal media for wastewater.
3.0
AQUACULTURE
The aquaculture industry in Idaho annually grosses over $80 million. This industry is
primarily located in southern Idaho over the Snake River Plain, a sole-source aquifer. The principal source
of water for this industry is the thousand springs area of the Snake Plain Aquifer (Figure 1). It is estimated
that this industry uses 2000-3000 cfs of ground water. The successful and profitable raising of fish
necessitates high-quality water which is at least as good as drinking water and in some cases must be more
stringent. Figure 1 is a generalized flow map of the Snake Plain Aquifer showing ground water flow
directions and land use activities. This is a typical agricultural area with associated row crop irrigation
practices, pesticide and fertilizer application, dairies, feedlots, food processing industries, and small urban
areas. As can be seen, the $80 million a year aquaculture industry is very dependent on successful
management of the Snake Plain Aquifer and multiple sources of contamination.
Generalized Cross Section of
The Snake River Plain Aquifer
INEL
Agriculture
, Dairies
Feed Lots.
Food Processing
Land
Aquaculture
, Groundwater Flow
Snake Plain Aquifer
4.0
SURFACE WATER RECHARGE
In some areas of the Western United States, ground water forms over 80% of the base flow
for streams and lakes. Consequently, degraded ground water can significantly impact the surface water it is
recharging. One of the most significant nonpoint source pollution problems in the United States is caused in
part by degraded ground water. The Snake Plain Aquifer annually discharges over six million acre-feet of
water to the Snake River in the thousand springs area. This ground water is contaminated with nitrates
which measures less than the MCL of 10 mg/1 for nitrates. This ground water then combines with
phosphorous loading from surface water to cause algae and macrophyte growth of unprecedented
GW3-20
-------
Session GW-3
Jehn
proportions in the thousand springs area of the Mid-Snake River in Idaho. The "floating" rafts of algae and
macrophytes inhibit boating, recreation, fishing, and cause increased restrictions on various permits mandated
under the Clean Water Act (i.e., NPDES, 401, 404).
This situation is caused by a cumulative impact on the surface and ground water by all
activities in the area including: septic systems, agriculture, aquaculture, feedlots, dairies, industry, land
application of wastewater, stormwater, and municipal discharge.
Figure 2 is a generalized cross-section of the Snake Plain Aquifer showing various land use
activities.
Groundwater Protection URS
Land Use Activities
Drinking
Water
Standard
Non-Degradation
Non-Degradation
Agriculture * Septic * OaiiiM + Feedktt * Land Application » Food Proces«ng + Aquaculture
ActivntaM System) of Wane Water
Groundwater Flow Direction
Increased Land use Activities
A. In the upper areas of the aquifer, hundreds of miles from the discharge point to the
Snake River, is the Idaho National Engineering Laboratory. It was at one time
thought that the remoteness of the area and the vast distances to ground water
(over 500 feet) precluded any serious concern about ground water contamination.
Casual disposal practices are now known to cause serious ground water
contamination problems.
B. This region is mostly desert, with little water usage.
C. This region is mostly agriculture. The main uses of ground water are irrigation
practices and drinking water.
D. In this region, in addition to drinking water and irrigation practices are added food
processing, land application of wastewater, feedlots, and dairies.
E. The main uses of ground water in this region are drinking water, aquaculture, and
as recharge for the Snake River, through an extensive number of springs
discharging into the river.
The most restrictive uses of ground water, (i.e., drinking water, aquaculture and surface
water recharge) are the end use of the Snake Plain Aquifer. The cumulative impacts of land use activities
have and are continuing to impact down-gradient users of ground water.
GW3-21
-------
Session GW-3
Jehn
5.0 MINING/MINERAL PROCESSING
The mining industry can impact ground water quality through acid mine drainage, tailings
impoundments, milling operations, and through underground mining. Frequently, these same operations
require uncontaminated ground water for successful mineral extraction.
A gold mine in Washington annually produces over 40,000 ounces of gold (annual gross is
approximately $14,000,000). The mill uses flotation and cyanidation for mineral extraction. Mill water
cannot be reused because contaminants which accumulate in the milling process reduce the efficiency of the
milling operation and therefore reduce the amount of gold that is recovered from the ore. The mill water is
pumped to a tailings impoundment. Fresh water is piped over three miles to the mine site.
6.0 GROUND WATER AS A DISPOSAL MEDIA FOR WASTEWATER
Like surface water, ground water is used as a disposal media for wastewater. Septic
systems, injection wells, agriculture, urban runoff and land application of wastewater from: food processing,
dairies, sludge, and irrigation tail water, all have the potential to, and frequently do, impact ground water.
Two different management policies are typically considered to find the most effective balance between land
use activities and protection of ground water. The two policy approaches are nondegradation and
antidegradation. Nondegradation means that ground water cannot be degraded below the level of quality
found before any land use activities or contaminants impacted the ground water. A ground water quality
nondegradation policy would essentially mean protection through no land use activity. An antidegradation
approach means that the ground water quality must be maintained to protect existing uses or that ground
water can be degraded to drinking water standards. The exception to this is ground water with exceptional
wildlife and ecological significance which must be protected at their existing high quality.
These activities requiring disposal are just as critical to human daily routine as consumptive
uses of ground water. A balance must be achieved between disposing of our collective waste and protecting
ground water. A hard-line nondegradation approach would essentially mean that approving no more land use
activities would be used as the means to protect ground water. With this in mind, an antidegradation
approach has been adopted by Idaho in the Ground Water Quality Plan.
This does not imply that ground water should be used as a media to dilute hazardous
wastes. A proper use of ground water is as a receiving media for wastewater generated through responsible
land use management. For example, hi North Idaho, the Rathdrum Prairie Sole Source Aquifer was being
contaminated by nitrates from septic systems. County ordinance now requires that five acres is the minimum
land required for a septic system. It has been found that the cumulative effects of one septic system in five
acres does not cause nitrate degradation of the aquifer. The result here is a balance of protection and
human activity.
Over the past 20 years or more, ground water has been degraded by activities designed to
protect and clean up surface water. Permits issued under the National Pollution Discharge Elimination
System (NPDES), mandated under the federal Clean Water Act, have degraded ground water. This
degraded ground water is recharging and polluting the surface water the NPDES permit was written to
protect. Under NPDES permits, water was discharged to a holding pond which percolates the waste to
ground water. Unfortunately, this ground water is hydrologically connected to surface water in most areas,
which results in contamination for both resources.
7.0 DRINKING WATER AS A USE OF GROUND WATER
In Idaho over 90% of drinking water is supplied by ground water. Contamination of
drinking water by organic solvents, pesticides, chemicals, and bacteria is an increasingly apparent problem.
For example, 25% of the drinking water systems in Idaho serving over 3,300 people show contamination by
VOCs. Testing for contaminants under the Safe Drinking Water Act is expensive and can be cost prohibitive
GW3-22
-------
Session GW-3
Jehn
for small water supply systems. Through effective ground water prevention activities, the Idaho DEQ
estimates that over $500,000 can be saved each year in analysis costs alone because some systems will then
be shown to be "not at risk to contamination."
Increased pumping by new drinking water wells and ground water remediation wells has
caused changes in ground water flow direction. In some cases this change has inadvertently impacted
another beneficial use. For example, In North Idaho, a drinking water well was taken off line because of
contamination by organic solvents. This water system was connected to a larger "contaminant free" water
system. A new well was drilled for the larger water system at a cost of $500,000 to increase production.
This new well then captured the contamination plume. The original contamination incident is now
threatening a water supply serving 25,000 people. The $500,000 well was taken off line and is now only used
for fire protection back up.
In Boise, residents of a mobile home park have unknowingly been drinking water
contaminated with tetrachloroethylene for years. Due to this contamination, the Idaho DEQ ordered this
small water system to either treat the water for contamination or find an alternate water source. Fortunately,
an alternate source of water was located. However, this required the residents to pay a "hook up" fee to the
new water purveyor and higher monthly water bills. This mobile home park, serving less than 100
individuals, is primarily used by limited-income and lower-income individuals who can least afford the
additional costs. The source of contamination is thought to be at least 1 mile upgradient and is not
connected to any current or past activities of the mobile home park.
Current regulations under the SDWA define a public water supply as 15 connections or 25
people served. In Idaho, about 30% of the population receives water from water supplies which are not
regulated under the SDWA. This Includes private wells and water systems which serve less than 10
connections or less than 25 people. These systems are not required to test for contaminants. Frequently,
these small systems are more vulnerable to contamination than the larger systems, due to drilling wells in
shallow ground water, improper well construction, and lack of routine maintenance. The long-term health
impact on individuals drinking water from these sources is unknown.
8.0 DISCUSSION
The nature of ground water, laminar flow, and low velocity combined with the lack of
biologic activity and the absence of ultraviolet radiation, inhibits the natural dilution or degradation of
contaminants. Even small contamination incidents, (i.e. less than 10 gallons of dry cleaning solvent) can
render billions of gallons of water useless. One example of this is the municipal water supply system serving
Garden City, Idaho. Current testing Indicates that the Garden City water supply system has been
contaminated by the cumulative impacts of household and small business practices for hazardous waste
disposal over many years. Once degraded, ground water can remain unusable or even hazardous for
hundreds of years. Thus the cumulative impacts of land use activities can be devastating.
It is estimated that only a small percentage of ground water has been contaminated.
Frequently, this is the same water which is being withdrawn for multiple beneficial uses.
The solution: a nondegradation policy for ground water is impossible; as it is for surface
water. An antidegradation policy is achievable. In this management scheme, land use activity and the use of
ground water as a resource is managed.
Under an antidegradation approach, the use of septic systems and the land application of
wastewater would be possible if no contaminants could reach the ground water. However, all land use
activity, not just septic systems and land application of wastewater, would have to be severely curtailed if
these contaminants reach the ground water and meet the MCL for drinking water.
GW3-23
-------
Session GW-3
Jehn
Land application permits have a point of compliance where ground water quality is
measured for compliance with drinking water standards. This point of compliance is sometimes the property
line. Depending on the current and future beneficial uses of the aquifer, a value of 10 mg/1 nitrate, at this
point, may or may not be appropriate.
The identification of future beneficial uses is frequently based on best professional
judgement. In Florida, there are aquifers which are now being used for drinking water that, as recently as 20
years ago, were considered nonpotable due to high total dissolved solids. Unanticipated population growth
has demanded that this water be utilized for drinking water.
Drought, climatic changes, population growth, disaster are all too variable to effectively
predict 100% of the future uses of ground water. Given the fact that, in many cases, ground water is
technologically unfeasible to remediate and that, once contaminated, ground water can take hundreds of
years for "natural remediation" to clean itself, it seems prudent to protect ground water for the more
restrictive beneficial uses.
The liability of contaminated ground water can be tremendous. The condition of
contaminated ground water is frequently considered during property assessment. Property values have been
degraded because the ground water is no longer suitable for drinking. This has occurred through actions by
current owner, past owners (unknown to the buyer at the time of purchase), and from migration of
contaminants from upgradient sources. Areas with or near contaminated ground water are not economically
attractive to developers.
EXAMPLES OF GROUNDWATER VALUATION
Aqua Culture
$80,000,000 per year
Mineral Processing For One Gold Mine
$40,000 oz gold per year X $350 per oz = $14,000,000
Drinking Water
1,000,000 people X $15 a month X 12 months = 170 million
Food Processing
Three Potato Processing Plants in Idaho
use 7 Million Gallons of Untreated Groundwater per day
Surface Water Recharge
How Much is Water Recreation Worth?
GW3-24
-------
Session GW-3
Jehu
9.0 CONCLUSIONS
Numerous examples and calculations can be performed to demonstrate the value of ground
water or the consequences of contamination. In many cases, millions of dollars of net worth can be assigned
for a localized beneficial use of ground water. Attempts to assign a value to ground water are as futile as
are attempts to assign a value for surface water. Ground water, like surface water, must be managed as a
natural resource and protected for identified beneficial uses.
Surface water beneficial uses include drinking water, agriculture, aquaculture, recreation,
and as a receiving media for point and non-point source contaminant discharges. Ground water has similar
beneficial uses. Ground water and surface water must be managed jointly because they are one media
through hydrologic interconnections. What affects one ultimately impacts the other resource.
The uncertainty of future demands or needs for ground water necessitates the need for
managing ground water as a natural resource. The concept of nondegradation is too restrictive and the
concept of managing ground water only for "reasonably expected uses of drinking water" is erroneous.
Prevention of contamination is clearly less expensive than remediation or treatment. Even
one contamination incident can cost hundreds of thousands of dollars in remediation, or treatment costs.
GW3-25
-------
Session GW-3
Cole
Policy and Regulatory Considerations
(transcribed from audio tape)
Keith Cole
Minority Counsel to the Subcommittee on Oversight and Investigations
Committee on Energy and Commerce
U.S. House of Representatives
I was writing this out yesterday and making a few notes after this morning's presentation. I
am afraid I am going to come off as fairly cynical I do not mean to belittle in any way the efforts that have
gone into the studies we have heard about this morning. I guess my concern is that they get chewed up in
the political process and are not thought of in the same way, at least I have heard people talking about how
they viewed these studies in this morning's session.
We are asking the question "What is the value of ground water?" It is not really designed to
produce a specific answer. It is more designed to restructure your world view a little bit and produce an
understanding of the world around you more than it is designed to come up with a specific answer.
What I heard this morning were presentations that at least appeared to come from the view
that this question, what is the value of ground water, is something that we can go out and do research on
into avoidance costs, into contingent valuation, nonuse costs and we can find what the value of ground water
is. It is something out there that is discernable. I would submit that the way the political process views this
is quite different, and I would submit that the studies we have heard about this morning cannot answer the
question and while these discussions, these studies, are not irrelevant to how the Hill looks at the question,
they are relevant only in a secondary order, and that is what I want to get at today and maybe draw a few
conclusions from how they are relevant.
I am not an economist, so I am sure a lot of this morning went over my head and I will be
the first to admit my lack of a grounding in economic theory. But very few Congressmen are economists
either and one of the things I have learned from working for Congress is it works off of very simple
analogies. It works off of anecdotes and things that are easy to visualize. My sense is that the question
about what is the value of ground water is a lot like the question of how much is my house worth. If I go
out and ask the experts, they will do studies, they will do some type of expert appraisal, and they will come in
and they will compare houses around the block and over tune and different sizes, and they will give me a
number or series of numbers, but hi the end, from my perspective as a home owner, the value of the house
is what someone is going to pay to buy it and those statements about what the house is worth, what its value
is based upon appraisals are interesting, and they might tell me something about did I get a good deal or
not, they might make me feel better or worse. They might make me feel I did not get as much as I should
have or I got more than I should have but, in terms of what the value of that house is, it is what I get paid
for it.
I would propose this thesis to the answer as to what the valuation of ground water is. The
valuation of ground water is essentially equal to the resources directed by Congress, the states, the
implementing agencies, and private actions for avoidance or remediation: the amount of those resources
devoted to its protection or remediation.
And so, in the end, it is Congress and the states, private individuals, and companies that will
ask the question what is the value, and what we heard this morning are really normative statements about
what some person or some group of people believes the valuation of ground water ought to be. There is no
value out there that we can assign to it. There are just competing groups; there is nothing scientifically
discernible and it is all up to politics. As an engineer, I wish there were some objective method by which to
GW3-26
-------
Session GW-3
Cole
arrive at an objective number and then, from that, we can agree on what this price is, and from that we can
agree on some level of expenditures to protect it.
However, I am forced to conclude that the value of something like ground water is
fundamentally a political decision and not an economic one and not a scientific one. I get there from a
couple of years of sometimes frustrating experience in dealing with Congress and how it reacts to factually
difficult questions and complex scientific issues. I bring this up, not because I am trying to convince you of
my reading of what the value of something is. I am sure the economists in the audience will critique me for
a lack of full understanding of it, but I think something flows from this perspective as to the grounding for
the valuation of ground water, no pun intended.
Suppose we take it from this perspective, the value of ground water is not defined as
something scientifically observable, but it is something that is defined in the end by the political process.
Studies cannot tell us what the value of ground water is, nor can they say really whether a study has
produced a better number than another one, but they are not relevant. They are not relevant because these
studies will be used as rhetorical weapons in Congress, in the states and amongst the implementing agencies.
I guess my cynicism comes in when I say that I despair as to how they will be used.
Let me give you an example. EPA views itself in many ways as a conservator and advocate
for the environment and we can expect, as an institution, EPA will be out searching for ways to find higher
value for ground water. For example, one might describe OMB as a part of an institution within the federal
government that considers itself to be an advocate for the overall health of the economy. We can expect it
to search for either mechanisms that appear to demonstrate lower value to ground water or higher costs.
Negative outcomes that flow from decreased GNP based on regulations would be one such new discovery
that we have up there.
We have EPA discovering new values and OMB discovering new costs. We also have a
similar dynamic in Congress where advocates for the environment will discover new values to ground water
and use your studies in this way, advocates for a different perspective will discover new costs, discover less
value. Unfortunately, the terms have been simplified to the point of nonrecognition and all your fears about
aggregation will be realized with a vengeance on this one. If an economist contemplating the use of his
studies will be well advised, some wit quipped that if Martin Luther even imagined Unitarianism he would
have cut off his right arm rather than nailed his principles to the door.
In the midst of all this frustration with how the issues that we described in the morning will
get ground up in the political process, is there anything positive to say about the impact of these studies on
the political process? The perspective that I tried to show to this point is one in which the valuation of
ground water is self-referential in a way. One of the speakers this morning mentioned that valuation is a
function of the perceived fear of cancer; one of the highest correlations with this data was this perception of
the fear of cancer resulting from this.
This is, in turn, a function that perception is, in turn, a function of media coverage, of
political energy being focused on a specific types of problems. So, in a way, what we have is a self-referential
system, can we use the value of ground water to drive the political process? What we see this morning is
that this self-referential nature of the political process in turn helps to define the value of ground water.
My conclusion is that you cannot speak of the value of ground water. What you can speak
of are many values, something that was hinted at in the discussions of the aggregation problem. If we have
many values of ground water, I do not think we can search for a number that we can plug into some formula
of cost benefit analysis at OMB. I think that is chasing a shadow. What we have to confront is that the
answer is going to be many things to many people. The value of ground water is going to vary by location.
It could be the geography or demographics of the locale. It could vary within the location by income
distribution. It is going to vary over time, and with media exposure and political energy directed to whether
risks both by ground water are greater or less.
GW3-27
-------
Session GW-3
Cole
distribution. It is going to vary over time, and with media exposure and political energy directed to whether
risks both by ground water are greater or less.
I do not think we have seen anything this morning that leads me to conclude that I have a
metric for judging whether any one of these valuations is right or wrong. What I see is a cacophony of
voices making different statements about value. This mess is very reminiscent of Congress where you have
435 representatives in the House and you will have as great or greater a spectrum of statements on value
within the Congress of the United States as you would surveying the public.
Let me just make a couple of conclusions about the flow of how these studies will be used
in Congress, and how they do not reflect the self-reverential nature of this, and the multiplicity of voices and
multiplicity of values that we will be able to find out there.
First, decentralize the ground water valuating decisions. Retain this at the local level. That
is a message that we have amongst the many conflicting and ambiguous Congressional statements. To date,
that has been one that has been fairly clear. Recognize that and allow for differences in ground water value.
Create the system based upon a belief that you are not trying to look for a particular value but you will look
for and allow for a multiplicity of values. Lastly, bear in mind that the technology that we have today will
not give us clean ground water. Once this becomes contaminated, we do not have the technology to
completely remediate it. I would look to distinguish and provide separate values for preservation of ground
water and values for ground water that is to be remediated.
Finally, any study that is based upon information that the public is asked to assume that
something is to be 100% effective or completely safe is in my mind rather suspect, particularly in the
remediation phase, because of the technological failing to get 100% remediation and also in light of the
schisms over what people view as safe levels.
GW3-28
-------
Session GW-3
Powell
Policy and Regulatory Considerations
(transcribed from audio tape)
Jimmie Powell
Staff Committee on Environment and Public Works
U.S. Senate
It is a pleasure to be here. Our committee, the Committee on Environment and Public
Works in the Senate, has jurisdiction over the broad range of environmental laws. In the most recent
Congress that just adjourned, we spent most of our time working on the Resource Conservation and
Recovery Act (RCRA) and solid waste disposal law and reauthorizing that law.
I thought I would start today by reading you just one little piece of the bill that the
Committee on Environment and Public Works reported. It was an amendment offered by Senator
Durenberger. It was a short amendment. It deals with ground water policy. RCRA is principally a ground
water protection law but you would not know it by reading the law. There is not much reference to ground
water. It is mostly a reference to wastes. There is no place in RCRA currently that expresses the value that
the country places on ground water, so Senator Durenberger tried to fill that gap with the following
language.
"The Congress hereby declares it to be the national policy of the United
States that the quality and quantity of ground water resources of the nation
are to be protected for the widest range of uses, both for this generation
and for posterity. Although some minor portion of the ground water
resources of the nation may be irreversibly or irretrievably committed to
use and consumed during the stewardship of this generation, it is contrary
to the policy of the United States to allocate to future Americans the cost
of ground water contamination associated with current social and economic
activities. For purposes of this act, [this it the important part] protecting
human health and the environment includes the prevention of ground water
contamination and the correction of contamination which has already
occurred."
That language was adopted by the committee after an hour of debate that was intense and
informed, I thought, more than debate usually is in the United States Senate. It expresses a pretty strong
view of the value of ground water resources and the nature of the value system that the members chose to
bring to the question in RCRA. The principal value system is protection of the resource and choices for
future generations.
Senator Durenberger first offered that language as a bill in response to the debate in the
Administration on the municipal landfill rules that were being promulgated under Subtitle D of RCRA. As
you recall, in October of 1991, EPA put out new Subtitle D of landfill criteria under amendments that were
passed in 1984 and led up to that promulgated rule. There was a debate about whether or not liners were
going to be required, how much ground water monitoring was going to be required, whether small
communities would be able to opt out of certain of those provisions, whether in the West, where the rainfall
was less and the climate was different and the depth of the water table was much farther, that requirements
would be less stringent.
One part of the administration, at least, took the position that it was only necessary to
protect the current users of ground water, that in fact, unless you could show a direct connection between
the hydrology of the landfill and a well nearby that was being used for human consumption, it was not
necessary to worry about ground water contamination, and only in those few cases where you were talking
GW3-29
-------
Session GW-3
Powell
about real people using real drinking water supplies that were near landfills was it necessary to worry about
ground water monitoring and liners in leachate collection systems.
That was a view that was strenuously opposed on the Hill, on our side at least. Many
letters went down urging that the provisions of the rule be more stringent and that liners and the ground
water monitoring be required.
Eventually, the rule was improved. Other parts of OMB, actually Bob Grady, I understand,
got involved in the decision and was persuaded by the cost of cleanup, looking at the future costs of ground
water contamination that might result in landfills that were inadequately protected. Mr. Grady asked that
new numbers be run on the costs of the prevention measures to see how they compared to the future
cleanup costs, and to the extent that prevention measures today were justified by future uses of the ground
water supply, he considered it appropriate, and the rule came back in the direction of the original proposal
and the final promulgated rule was somewhat more stringent than the intermediate proposal.
All of that is by way of introducing three outlooks on ground water that you can sometimes
hear if you are listening on Capitol Hill. The first outlook is the protect the current user outlook. It leads
to classification systems which EPA has proposed. Look at who is using it in industrial use, or agriculture
use, or drinking water. It led to the state ground water strategy that the agency recently proposed that left
most of those decisions at state level and promised that, notwithstanding the legal requirements of federal
law, that you would be able to do it best to respect state decisions.
It leads to documents like the Unfinished Business report from the agency that put very low
priority on programs such as Superfund and hazardous waste and programs under RCRA, because the media
being protected was ground water, and much higher priority on air pollution, indoor air pollution, and
worker exposure and global issues because the exposure was much higher.
So that is the current user's notion and I think it is fair to say that it is a view, an outlook,
that you will hear most often from the administration, most often in the context of RCRA-prevention
measures.
The second outlook is the cost of cleanup, the view that Mr. Grady expressed, and I think
the best representation of the "protect the current user's view" is the cost of cleanup. Mr. Grady must
understand the expense that the military part of the government is facing as it prepares to clean up the 19
weapons production facilities across the country. All of that is ground water contamination. The federal
government is facing a $120 billion cleanup cost. No one is proposing that, no one is using it right now, so
why do we get to clean it up? The federal government, DOD, and the DOE understand that they have to
proceed as rapidly as they can, as resources allow, to clean up those facilities. And it is that huge expense
which, I think, educates one that it is not just current users, but it is also the expectation that the resource
will be passed along in good shape that drives government policy.
The third outlook is called nondegradation. It is a view that some states have adopted. The
state of Minnesota which Senator Durenberger represents, is a nondegradation state. Wisconsin and some
other states have non-degradation. The view is that, wherever you possibly can, you prevent ground water
contamination, and, where you discover contamination, you take corrective action so you put the resource
back in as good a condition as possible with current technology.
The first two outlooks, the protector and users of the cost of cleanup outlook, accept the
notion that a certain amount of contamination is inevitable or acceptable, and that it is a matter of managing
your activities so that the amount of contamination does not get out of hand or go beyond the cost of
cleanup or affects current users. The non-degradation view does not accept the notion that there is any
acceptable contamination.
GW3-30
-------
Session GW-3
Powell
I think there are a series of principles that you can apply in the case of ground water that
lead you to a nondegradation view. Analyses of the situation we currently face, and that judgment about
whether or not it is reasonable, in the context of that analysis, to continue to maintain a view that you can
have a certain amount of contamination that is acceptable. One is the behavior of ground water itself; the
connection between the activity on the surface and the quality of the aquifer usually is poorly understood,
and complex Years and years separate the event and the outcome. Farmers apply pesticides or fertilizer to
their fields. They do not actually contaminate their water table for quite some time; by the time they
discover the contamination, changing their practices on the surface usually will not have the preventive effect
that they were hoping.
The second problem is the effect of the contaminants themselves. We often pretend that we
fully understand the health effects of these contaminants or environmental effects in that we can create these
elegant mathematical models getting from clinical data on animals and cancer risks to some sort of low-dose
human exposure, but all of that is tenuous, as all of you know, and it also leaves out all of the other effects
that we so poorly understand. The other chronic effects, the birth defects and neurological effects we are
only beginning to appreciate. The second problem with the first two outlooks that say you can manage
contamination is the effect of the contaminant.
The third problem is risk. Americans are much more willing to accept small risks spread
evenly over the whole population than they are large risks that are imposed on the few, especially if those
large risks are involuntary, imposed by industrial activities that they did not benefit from. Ground water that
is contaminated at high levels, and the people that experience that contamination experience very high risks.
The models that we use today that take a flat-risk, linear-risk curve and assume that, just because exposure is
high, we are necessarily dealing with a more important public health threat, do not reflect the value system
that the American public has.
The final problem with that approach is the value that people put on the future in the
United States. It is not simply a discount function. It is entirely a jump function. It is different. It is
disproportional.
The best analogy for ground water contamination is the budget deficit. When Senator
Durenberger presented his amendment to the committee and urged the committee to adopt it, he talked
about passing on ground water contamination being equivalent to passing on a budget deficit. And that
really resonated with the members. Members understood that and they came right back on that point. The
people who spoke in favor of the amendment spoke in those terms.
It is a decision that we do not know enough to manage a resource and hand it on in a
condition that is less that the condition we received it and be sure that we are allowing the next generation
the choice that is as good as the choice that we received.
If you worked in the Senate or the House, I think you would discover, as Keith indicated,
that the level of discussion certainly is not as sophisticated as it is here today, and that it is the analogy or
the anecdote or the more general statement that drives the legislative process. During the period I have
worked for the Environment and Public Works Committee, we have had three different members chair the
subcommittee with jurisdiction over ground water, Senator Durenberger, Senator Backus, and now Senator
Moynihan. With respect for all three members, it is very difficult to get up the education curve on ground
water. It is a hard resource to understand. It is different than air and surface water and it does not mix. It
moves slowly. Once it is polluted, it is hard to remediate.
All of those are aspects of ground water that make it difficult to understand. It is a huge
victory just to have a United States Senator finally get his head around the resource and understand how the
resource is different and how it operates as a resource, as a medium of carrying contamination. And when
you get there, they all jump to the nondegradation conclusion. Once you are educated to the nature of the
GW3-31
-------
Session GW-3
Powell
resource, their value system says that is just like the budget deficit and we are not going to pass
contamination onto future generations.
And in fact, and this my concluding thought. They are more willing to pass on a budget
deficit than they are willing to pass on ground water contamination. If you gave them the choice of saving
the money on the military bases, the $120 billion, and passing on the contamination, or spend the money now
and passing on the deficit, they are going to choose passing on the deficit, I am fairly certain. They are big
dollars and they are prepared to spend them.
Senator Durenberger introduced a bill hi 1988, the Comprehensive Ground Water
Protection Bill, that attempted to take the nondegradation, future generation view, and apply it sort of in a
generic context to all the sources of ground water contamination from pesticides to cemeteries. The bottom
line on that piece of legislation hi that effort was prevention means using the best available technology in
practice for each of the principal sources of contamination, that nondegradation is not giving up and quitting.
But it is, as we do in so many other laws, doing our best to control the discharges so the
ground water resource is not contaminated and then, when contamination does occur, doing the very best
with available technology to clean it up. It is not a claim that the resources must remain pristine forever.
We use coal, oil, natural gas, and whatnot, and they use the ground water resources just like those; it is just a
plea that we not inadvertently pass on contamination as the cost of our activity to future generations where it
can be avoided.
It is unlikely that the Congress will ever enact a comprehensive ground water protection law,
but I think it is to be expected that, as we move through RCRA and clean water, we will address many of
the sources of ground water contamination that are not yet addressed completely, and that some version of
best available technology and best practice will be applied, and it will be applied because the Congress' value
for ground water is a nondegradation value, motivated by a generational value system and the doubt about
our ability to manage a resource like ground water in a way that finds some level of contamination
acceptable.
GW3-32
-------
Session GW-3
Jacobson
Policy and Regulatory Considerations
Debra Jacobson
Counsel to the House Energy and Commerce Subcommittee on Oversight and Investigations
U.S. House of Representatives
The scene that my two Congressional colleagues have emphasized is one that I am going to
repeat with a slightly different twist, tying in hearings that the Subcommittee on Oversight and Investigations
has held. The subcommittee has held hearings during the past two years, which have focused in the ground
water area on the implementation of Executive Order 12291. This Executive Order issued by President
Reagan, states that regulatory action will not be undertaken unless the potential benefits to society for the
regulation outweigh the potential costs to society.
The first subcommittee report in this area was in 1979 and it was entitled "Cost Benefit, or
Mirage." It critiqued the use of quantitative cost-benefit analysis as a decision-making tool. There is a view
that an analysis of costs and benefits can be helpful in educating the decision maker. However, concern has
been raised about using cost-benefit as a decision rule, requiring quantified benefits to exceed quantified
costs. There has been a consistent rejection of this view by the subcommittee in certain reports at hearings,
but there have been statements by various members expressing strong concern.
I think a lot of the reasons for this opposition have been discussed by my colleagues on the
panel here. There are a lot of values which Congress reflects, and they are not only values which can be
quantified in an economic sense. There are ethical values. Do we really want to impose the burden of a
chemical industry's operations on a few families that live around that plant? There is an ethical feeling
questioning whether these families or small groups of families should bear those costs even though
economists might be against regulatory controls because of the high cost per death averted.
There are also values in terms of preservation of a resource for future generations. Some
of those values are reflected in religious values. I think one of the reasons that there has been so much
focus on this is there really is a multitude of values which cannot be strictly quantified.
In terms of the subcommittee's efforts in this area, there were hearings held on ground
water issues in 1991, which focused on pesticides in ground water, but also touched upon the RCRA
Municipal Landfill rule. There was another hearing held in 1992 that focused in large part on the role of not
only the Office of Management and Budget but also the Vice President's Council on Competitiveness in
terms of ground water policy.
The landfill rule was focused on because OMB initially argued that the rule should not be
issued because the cost which had been quantified by EPA in its regulatory impact analysis exceeded the
benefits which had been quantified hi the regulatory impact analysis. EPA had not quantified the benefits of
preserving ground water for future generations; it had not included a discussion of equity issues. EPA had
felt that such analysis would take several years, and the rule already was several years past the deadline of
1988 that had been imposed by Congress for issuing the rule. OMB felt that liners should not be required in
landfills because the benefits were not there, they felt, to outweigh the costs of putting liners in landfills.
And so this was focused on by the subcommittee because it really was a case where cost/
benefit analysis was combine with quantitative risk assessment. There was an attempt, really, to use cost/
benefit analysis as a decision-making rule. There also was the view that this approach was inconsistent with
the intent of Congress expressed in the 1984 amendments to RCRA.
There has been a lot of discussion today about the view that ground water is a resource and
it should only be protected for current users and reasonably expected drinking water users That is an issue
GW3-33
-------
Session GW-3
Jacobson
that also has been pushed by the Office of Management and Budget and reflected in the debate over the
landfill rule. That is stated most precisely in the EPA's ground water strategy for the 1990s, which was
heavily influenced by the Office of Management and Budget. It stated that ground water should be protected
to ensure that the nation's currently used and reasonably expected drinking water supplies do not present
adverse health risks and are preserved for present and future generations.
The committee's hearings focused on trying to discern what did this term "reasonably
expected drinking water supplies" mean. Our subcommittee members expressed concern. Jimmy mentioned
several of these issues earlier. Could we reasonably have expected the condominium growth and heavy
ground water demands in Florida and Arizona 50 years ago? Could we reasonably expect that the
Stringfellow Acid Pits in California would be leaking and creating one of the most significant Superfund
sites? At the time that wastes were placed in that landfill, the view was there was impermeable bedrock
under that landfill and that it would never, ever leak. That certainly has not been the case. So there has
been a lot of concern expressed on Capitol Hill, and it was expressed here by some of the state
representatives, about the view that we really should be careful when we are talking about cutting back on
prevention, because there are so many uncertainties in terms of population growth, aquifer hydrology, and
many other factors.
Jimmy also touched upon another related issue, which is the concept first set forth in the
EPA's unfinished business report in 1987 and then in the report of the Science Advisory Board's Risk
Reduction and Strategies Committee in 1990 or 1991, that we should allocate the resources spent hi
environmental protection on the basis of a relative risk ranking. In an extension of this effort, OMB, in the
1992 budget, had a chapter which ranked 53 health and safety regulations issued by EPA and other federal
agencies on the basis of a projected cost per death avoided. The chapter emphasized that the cost
effectiveness of the actions varied over more than eight orders of magnitude from about $100,000 per
premature death averted to more than $5 trillion per premature death averted for regulating wood preserving
chemicals and hazardous waste.
Ground water protection fared very poorly under OMB's risk ranking because the potential
benefits of the rules had not been quantified in the framework of premature deaths averted, and instead
largely accrued in the form of preserving ground water for future uses. Under the OMB approach, if there
was no population currently living in the area, there was no exposure to contaminated ground water, there
were no health risks, and therefore, there were no resultant benefits to reducing ground water contamination.
The OMB approach also focused largely on cancer risks and it excluded consideration of many other risks,
such as potential immunotoxic or neurotoxic effects.
In an effort to get the debate more focused on this issue, Chairman Dinsell wrote a letter to
the co-chairmen of the EPA Science Advisory Board's Risk Reduction and Strategies Committee, Dr.
Raymond Loehr and Jonathan Lash, and asked them to comment on some of the issues that the budget
chapter had brought out. They sent a letter back to the committee that expressed concern about the OMB
approach and stated that it was troubling for several reasons, and I think that these are illuminating.
First, they said the proposal assumed a level of precision in the estimation of health risks
that the available data simply do not support. A recent report of the National Academy of Sciences on
Environmental Epidemiology has stressed that in the area of health risks, there has been very little resources
and effort devoted to understanding a lot of the key health risks for contaminated ground water, such as
immunotoxic and neurotoxic effects. We have barely scratched the surface. The mechanisms of cancer on
some of the other effects are really not very well known. So the Science Advisory Board co-chairmen were
really saying is when you take quantitative risk assessment which, is so uncertain, and then you combine that
with quantitative cost-benefit analysis, we are really only in the preliminary stages of starting to even try to
quantify some of these benefits, then it is kind of "garbage-in-garbage-out" result. That is my words, not
theirs.
GW3-34
-------
Session GW-3
Jacobson
They also stated that the OMB proposal appeared to be largely based on the risk of
premature death from cancer, and that one of the main thrusts of their report had been that EPA had
focused too long on health effects to the relative neglect of protecting ecosystems and also had focused
largely on cancer health effects to the relative exclusion of other health effects. Their last point related to
the issue of intergenerational equity, and they thought that equity really could not be considered by a risk
ranking approach, and it was a factor that needs to be considered.
At the subcommittee's hearing on this matter, there was a colloquy between Congressman
Roland from Georgia and Deputy EPA Administrator Habicht that I think focuses on some of these value
questions. Mr. Roland said, "Mr. Habicht, would you agree that many of EPA's statutory mandates such as
RCRA reflect not only risk considerations but also issues of ethics and equity, such as which generation
bears the burden of cleaning up contaminated ground water and whether a pesticide manufacturer or an
individual family bears the burden of the adverse impacts of pesticide contamination?" Mr. Habicht agreed
that yes, those values are values that are implicit in reading RCRA and that RCRA clearly is not based on a
cost-benefit kind of equation. He also went on to say that he felt that there needed to be greater study of
ground water valuation matters. In fact, the hearings that the subcommittee held, I think, were partially
involved in spurring the to EPA to look at some of these benefits issues more closely.
I will conclude by saying that, in terms of the subcommittee's oversight efforts, we are
carefully following the regulatory impact analysis for the corrective action rule that EPA is preparing. I
believe that some of the work done by the presenters, such as Mr. Schultze here today, has been funded by
EPA in preparing that regulatory impact analysis. The subcommittee will be interested in looking at the
strengths and weaknesses of the new analytical tools that are developed.
However, I will state, going back to what my colleague said, that I think that any studies
that are developed will be used in the debate, and industry will use the studies that have lower benefit values
and environmental groups will probably use the studies that have higher values. The bottom line, however, is
that Congress, just like the American people, is not only reflecting a monetary cost when they make a
decision. They are reflecting the values of their communities, their concern about the future, their religious
beliefs, and the like. So I think that as people go forward in their research in this area, I think they have to
realize the complexity of the issues.
GW3-35
-------
Session GW-3
Questions and Answers
Questions and Answers
(transcribed from audio tape)
QUESTION: [inaudible] these days and how do we set priorities. I really want to hear the
answer to this one.
I think the scope is large but it is not like we start with having done nothing. I think we
look at fixes to the laws that we have on the books to start with. I think as a policy we need to put a goal in
the Clean Water Act to protecting ground water, to add that.
But we can do things like, for example, under the Clean Water Act, if you are looking at
facilities that get an NPDES permit for discharge and you are looking solely at what is happening on that site
with regard to a piped discharge, begin to look as you issue those and as you re-issue old permits, go in and
take a look at what is happening on ground water and for the time being, use best available judgement, best
professional judgment on what can be done to protect ground water on the site.
For example, if you had, say, a wood preserver who had a treatment process with a
discharge, you go in and you write their permit so that you are not only dealing with the points where there
is discharge but you are looking at the drip pad placement, you are looking at the spill control on the site,
those sorts of things. So you piggyback on the existing regulatory programs you have.
You go into the Resource Conservation and Recovery Act and you pull out exemptions that
are grossly unrealistic, the waste associated with oil and gas drilling and production should not be classified
in such a way that it ends up out of the regulatory scheme entirely. Go in and make fixes to those.
I would say also you take the lead from some states who are doing an excellent job in
particular areas in terms of setting up models about how you deal with particular sources, be they mine
waste, be they mining, be they septic tanks, and you distribute that, and you get the states trying to focus on
obviously they have to do some things first, so they first should be focusing on what are their priority sources
in their state, which is not to say that Arkansas will never deal with chicken processors, that Virginia will
never deal with coal mining, that New Mexico will never deal with uranium mining. That we really have to
deal with the sources on a priority basis, let the states take some lead. But I think we can build a lot on the
programs that we already have.
QUESTION: Mr. Weisrock, several attendees noted that you had referred to the shell
aquifer under Amoco's Yorktown facility and noted that there was no beneficial use of that aquifer and yet
as questioners point out it is right next to a river, isn't there a ground water/surface water interaction, and
isn't this a beneficial use?
DR. WEISROCK: Let me try to clear something up on this. Either I confused you or else
you got the wrong impression. The ground water underneath that refinery to the best of my knowledge is
not really an aquifer. In other words, the shallow ground water is not such that it could be delivered in
economically useable quantities. It is basically just water table and so, on that basis, my conclusion was that,
since you could not take it out of the ground in any economically useable rates or amounts, that, therefore, it
did not really have any beneficial uses.
I would have to agree at the same time that any time you have ground water and surface
water interconnected to where there is a link, you have to consider what the beneficial uses are under those
particular, in that particular case, and, actually, it is kind of interesting with the Yorktown Refinery. It sits
on the York River which empties into the Chesapeake Bay and the salinity of the water is somewhat high
because of tidal influences and so forth, and the refinery takes that water from the river, uses it as once
through cooling water, and then disposes of it since it is not, does not come in contact with any
GW3-36
-------
Session GW-3
Questions and Answers
contaminants, disposes of it in basically open trenches or ditches on the refinery property where it then
subsequently re-infiltrates into the ground water.
That is a perfectly allowable use, but, in that case, since the river water is somewhat saline,
we are actually contributing to the salinity of the ground water because of the impact of the river, and not
because of any impact to our operations. So I hope that answers that question.
The other thing I would like to do very briefly is just to kind of expand on something that
Thelma was saying a minute ago, because I think it is perhaps a good example of what she is talking about.
And that is, in a number of states, and I know Tennessee in particular has very recently made it a
requirement to get reissue of an NPDES permit for facilities such as a marketing terminal, that you show as
part of the basis for that permit, reissuing that permit, that you have, in fact, in place a ground water
protection plan which includes monitoring of ground water at the facility and that is something that we
definitely support, because, in terms of our ground water protection program, one of the things we want our
facilities to do is to develop these facility-based ground water protection plants.
QUESTION: Mr. Goodrich, why you did not talk about conservation and reductions in
water use as part of the solution in California and wondered why. Is this because recharge might be
adequate to compensate for the growth in demand and the drought?
MR. GOODRICH: In the interests of time, I sort of skipped over that part in my
discussion, but conservation is alive and well in California, particularly Southern California, which I am
closest to as far as the purveyors of retailers of ground water and surface water, is that it is very standard
procedure now to have ultra-low flow toilets, low flow showers installed in your home and in a new home.
There are programs for rebates to retrofit older homes which are very lucrative; for example, $100 rebate on
ultra-low flow toilets that use a mere fraction of the water of a standard toilet. The cost of those if $65, $75,
$80. So you walk away with a profit.
In Northern California, Central California, conservation in the central valley, a lot of farms
are lying fallow which relieves, makes water available to both people and the environment, and something
that is coming more to the fore and in the last several years is water that we did not necessarily set aside for
the environment is now a new demand and one that we are trying to face and allocate supplies for, so we
will be on the W word, weapons. We are building those weapons, we are designing them in Southern
California as artificial weapons and enhancing and stabilizing those in the Central California area.
As far as growth, that is sort of a separate issue. One thing in California, there is a recent
survey, I think it was off of the census, that showed that, I am going to make up a number but it is
something in that neighborhood, it is something like 1,700 people are leaving, adults are leaving California a
day, however, the major growth in our state is due to appropriation. We tend to have kids, 2.2 kids allotted
to us and that is enough to cause a growth and we have to deal with that. Personally, I am sending mine to
the East Coast.
GW3-37
-------
Session GW-4
Summary
Tuesday, October 20, 1992
Session GW-4: Issues and Research Needs in
Valuing Ground Water
SESSION SUMMARY
MODERATOR: Charles A. Job
PRESENTERS:
Marjorie M. Holland—Issues and Research Needs In Valuing Ground
Water: An Ecosystem Perspective
Robert Costanza—Issues and Research Needs in Valuing Ground Water
Eric J. Harmon—Liquid Assets and Paper Water: Valuation of Ground
Water Under Colorado's Prior-Appropriation System Traditional Basis
and New Issues
Stephen Crutchfield—Issues in Measurement of the Non-Use Value of
Ground Water
Maureen Cropper—Research Issues in Valuing Ground Water
Richard Howarth—Environmental Risks and Future Generations:
Criteria for Public Policy
GW4-1
-------
Session GW-4
Holland
Issues and Research Needs in Valuing Ground Water:
An Ecosystem Perspective
Marjorie M. Holland
Ecological Society of America
ABSTRACT
Members of the Ecological Society of America (ESA) recently set research priorities for the
discipline of ecology, and summarized these priorities in a report entitled "The Sustainable Biosphere
Initiative: Ecological Research Agenda." The topic of sustainable ecological systems was one of three
priority areas identified in this report. It is in the context of this report that we seek to understand further
ecological processes that determine the sustained functioning of natural and human-dominated ecosystems,
and to provide relevant information on the connectivity between river channels and ground water systems. In
many parts of the U.S.A, ground water systems determine the location of ecologically important wetlands.
Moreover, in healthy (higher valued) rivers, interstitial (subsurface) flow of water sustains unique aquatic
food chains that enhance water quality. Thus, a greater understanding of the connectivity between river and
ground water systems seems critical for improved management of our limited water resources.
TEXT
Many of the environmental problems that challenge human society are fundamentally
ecological in nature. The growing human population and its increasing use and misuse of resources are
exerting tremendous pressures on Earth's life support capacity. Humankind must now develop the
knowledge required to conserve and wisely manage Earth's resources. Ecological knowledge and
understanding are needed to detect and monitor changes, to evaluate consequences of a wide range of
human activities, and to plan for the management of sustainable natural and human-dominated ecological
systems.1
In response to national and international needs, The Ecological Society of American (ESA)
has developed the Sustainable Biosphere Initiative (SBI), a framework for the acquisition, dissemination, and
utilization of ecological knowledge which supports efforts to ensure the sustainability of the biosphere. The
SBI calls for: (1) basic research for the acquisition of ecological knowledge, (2) communication of that
knowledge to citizens, and (3) incorporation of that knowledge into policy and management decisions.1
The SBI report represents that culmination of a consensus-building process within the
ecological community to set research priorities. Three priority topics are identified in this report in the areas
of global change, biological diversity, and sustainable ecological systems.1 It is in the context of the SBI
report that we seek to understand further the ecological processes that determine the sustained functioning
of natural and human-dominated ecosystems, and to provide relevant information on the connectivity
between river channels and ground water systems.
In many parts of the U.S A., ground water systems determine the location of ecologically
important wetlands. Moreover, in healthy (higher valued) rivers, interstitial (subsurface) flow of water
sustains unique aquatic food chains that enhance water quality. Because most large rivers have extensive
floodplain aquifers which are hydrologically connected to the channel, it is important to evaluate the impacts
of altered flow regimes on ground water and riverine ecology. Instream flow studies have historically focused
on the physical habitat of fisheries and recreational values of flow and not on the ecological health of the
river. Recent research conducted in the Flathead River, Montana indicates that biodiversity in gravel-bed
rivers (Figure 1) is in large measure related to the existence of hypogean (underground) food webs.0-3* Ward
and Stanford found that ground water fauna (hyporheos) consist of two major elements: (1) members of the
stream benthos that temporarily move some distance into the streambed substrate (amphibiont); and (2)
GW4-2
-------
Session GW-4
Holland
specialized ground water forms that rarely, if ever, occur in the surficial stream bed (stygoboint).(Z> The
ground water zone penetrated by amphibiotic organisms is referred to as hyporheic habitat and the biota
present are the hyporheos.4
The hyporheic zone is defined as an ecotone or transition between the surficial stream bed
and the true ground water habitat. In recent deliberations, scientists have defined an ecotone as a "zone of
transition between adjacent ecological systems, having a set of characteristics uniquely defined by space and
time scales and by the strength of the interactions between adjacent ecological systems."5 In a watershed
context, transfers of nutrients, sediment, and energy may occur across several surficial and lateral
boundaries,6 including transfers between ground water aquifers into soils and between aquifers and stream
beds (Figure 2). The ecotone concept7'8 has fostered greater understanding of the extreme importance and
potential predictive power related to transformations and fluxes of materials that occur within boundaries
between functionally interconnected patches that form the riverine landscape.9
Many of the hyporheos recently found in gravel-bed rivers are new to science and may be
very sensitive to environmental changes wrought by humans.3 Hyporheic zones can be extensive. The
hyporheic zone on the Flathead River in Montana averages 3 km wide and 10 m deep, whereas, the channel
(median flow) is about 50 m wide.2 Stanford and Ward estimate that there was about 0.3 km3 of hyporheic
habitat compared to 125,000 m3 of channel habitat (within their study area) and that standing crop biomass
could easily exceed benthic biomass.
The hyporheic zone serves as a refuge for the surface benthos, offering shelter from floods,
drought, and temperature extremes, and providing suitable and predictable conditions for immobile stages
such as eggs, pupae, and diapausing larvae. The hyporheic zone also offers protection from large predators
and contains a fauna! reserve capable of recolonizing surface benthos, should they be depleted by adverse
conditions.3
Thus, a greater understanding of the connectivity between river and ground water systems
seems critical for improved management of our limited water resources. Wise use of ground water systems
requires action on a broad scale, giving consideration to all factors affecting those systems, adjacent
floodplain wetlands, stream channels, and other components of the drainage basins of which they are a part.6
We recommend that optimal management of ground water systems include: (1) management based on a
long-term, whole-basin perspective; (2) coordination between state, regional, and federal agencies that have
management responsibilities within the same basin; (3) coordination of public and private land owners
through local river basin councils; and (4) development of consensus for basin-wide goals to protect and
restore the connectivity between riverine and ground water systems. Thus, there is a need for a landscape
perspective in managing ecological systems, as well as in conducting research.
GW4-3
-------
Session GW-4
Holland
Figure 1. Major Landscape Features of the Kalis pell Valley of the Flathead River, Montana, USA, showing
the three primary spatial dimensions (lateral, longitudinal or altitudinal, and vertical) which are dynamically
modeled through time (the fourth dimension) by fluvial processes. Biota may reside in all three spatial
dimensions: riparos (streamside or riparian), benthos (channel), hyporheos (river-influenced ground water),
and phreatos (true ground water). The hatched area is the varial zone, or the area of the channel that is
periodically dewatered as a consequence of the average amplitude of the discharge regime. Major channel
features include a run (A), riffle (B), and pool (C); Sd refers to sites of sediment deposition, and Se refers to
a major site of bank erosion. The heavy solid line is the thalweg, and broken lines conceptualize circulation
of water between benthic, hyporheic, and phreatic habitats.
upland
aerobic soils
anaerobic soils
bedrock
< > lateral boundary
T surficial boundary
Figure 2. Generalized Diagram Showing Ecotones Between Inland Wetlands and Adjacent Systems
GW4-4
-------
Session GW-4
Holland
REFERENCES
1. Lubchenco, J., et. al. "The Sustainable Biosphere Initiative: An Ecological Research
Agenda," Ecology. Vo. 72, No. 2, pp. 371-412. 1991.
2. Stanford, J A. and J.V. Ward. "The Hyporheic Habitat of River Ecosystems," Nature. Vol.
335, No. 6185, pp. 64-66.
3. Ward, J.V. and J A. Stanford. "Ground Water Animals of Alluvial River Systems: A
Potential Management Tool," Proceedings of Colorado Water Engineering and Management
Conference. Colorado Water Resources Research Institute, Fort Collins, Colorado, pp. 393-
399. 1989.
4. Stanford, JA. and B.K. Ellis. "Proposal on Relation Between Discharge and Distribution of
Hyporheic Habitats Within Selected Stream Segments in Glacier National Park, Montana,"
Submitted to the Water Resources Division, United States National Park Service, 10 pp.
July 12, 1990.
5. Holland, M.M. "SCOPE/MAB Technical Consultation on Landscape Boundaries: Report
of a SCOPE/MAB Workshop on Ecotones," Special Issue 17. Biology International. IBUS.
Paris, France, pp. 47-106. 1988.
6. Holland, M.M., D.F. Whigham, and B. Gopal. "The Characteristics of Wetland Economics,"
Naiman. R.J. and H. Decamps, editors. Land/Inland Water Ecotones. UNESCO's Man
and the Biosphere Book Series. The Parthenon Publishing Group. London, pp. 171-198.
1990.
7. Naiman, RJ. and H. Decamps. The Ecology and Management of Aquatic-Terrestrial
Ecotones. UNESCO, Paris, and Parthenon Publishing Group, Canforth, United Kingdom.
1990.
8. Holland, M.M., P.G. Risser, and RJ. Naiman. "Ecotones: The Role of Landscape
Boundaries in the Management and Restoration of Changing Environments," Symposium
sponsored by the U.S. Man and the Biosphere Program and The Ecological Society of
America. Chapman and Hall. New York. 145 pp. 1990.
9. Stanford, J A. and J.V. Ward. "Management of Aquatic Resources in Large Catchments:
Recognizing Interactions Between Ecosystem Connectivity and Environmental Disturbance,"
Watershed Management. Springer-Verlag, New York. pp. 9-124. 1992.
GW4-5
-------
Session GW-4
Costanza
Issues and Research in Valuing Ground Water
Robert Costanza
President
Chesapeake Biological Laboratory
Society for Ecological Economics
What we have been doing with ecological economics is really not trying to create a new
discipline so much as to try and create and sustain a dialogue between ecologists and economists and
mathematicians, sociologists and anthropologists and all of the disciplines that are relevant to solving
problems like ground water, like global warming, like a lot of the problems that we are now beginning to
realize are really interdependent problems and cannot be adequately taken apart and addressed separately.
I am going to focus today on a couple of projects that we are just now starting. They are
both funded by EPA and address the ground water valuation issue and also the scaling issue that Marjorie
was talking about. The reasons that we have to now start becoming so interested in these interconnections, I
think, are largely because of the size, the scale, of the economic enterprise. The economic part of the system
has gotten to be so large relative to the ecological life support system that is necessary for us to explicitly
acknowledge the limits of the ecological system.
It is now necessary because of these inter-connections and because of the stress that we are
now putting on the ecological life support system, that we get a better idea of what it is really doing for us
and how much those services are worth relative to the other parts of the economic system.
It represents a change in world view. We can no longer look at the world as two
independent pieces. Ecologists tend to want to study pristine ecological systems which really do not exist
anymore. Economists tend to study the economic part of the system. But we are beginning to realize, I
think, together as a group, both from the ecologist and the economist sides, that these two parts of the
system cannot be taken apart that way if you really want to get a good handle on the kinds of problems that
we are addressing.
Also, the regional perspective is a very important one. Once you start looking at whole
watersheds, you begin to confront these interconnections directly. If you arrange your study around pieces of
the landscape that involve both ecological systems and economic systems and their interconnections, the flow
of water through these systems is a primary force connecting them. Then you cannot avoid this
interconnection any longer. It forces the issue.
So a lot of what we have been doing and what we will be doing in the Institute has to do
with looking at whole systems, whole watersheds and landscapes, and trying to understand the
interconnections between the ecological and economic parts of those systems.
And that leads us to the area of landscape modeling. A lot of my background and training
was in understanding and modeling large-scale ecological systems at the landscape level. And that involves
combining some methods and techniques that are just now making this sort of modelling possible. We use
geographic information systems to look at the way resources are arranged on the landscape and how they
have changed, historically, with time. We use simulation techniques to look in detail at the processes of
transformation at each point in the landscape. You can think of it as a big grid or array of cells and in each
cell there is a simulation model that looks at, hi our case, the flows of water and nutrients and plant
productivity and animal productivity. Now what we are doing with the one project together with Nancy
Bochstael, Ivar Strand, and Walter Boyton is to try and apply these techniques to the issue of the valuation
of ecological services generally, but ground water is an important one of those.
GW4-6
-------
Session GW-4
Costanza
We are going to extend what we are doing to include both the ecological and economic
components and use this interactive modeling approach to try and get at the issues of valuation. I think
there has been frustration on several people's parts about basing the valuation of ground water completely on
preference-based, survey-based kinds of techniques, without adequate analysis of the physical and biological
interconnection.
Contingent valuation is applicable to some of the values of resources like ground water, but
there are other values that it does not, and probably cannot, pick up. And, so what we want to do is to look
in more detail at the actual interconnections between these systems. It begins with the models explicitly and
then being able to manipulate those models to get at the direct and indirect impacts of making changes in
how we develop the landscape and how we use water and other natural resources.
Just to give you a quick example: we have completed these sorts of landscape models for a
couple of areas and are working on a few others. The one that is the most complete is in the coast of
Louisiana, where we were looking at the impacts of climate change, the impacts of river diversion, of canal
dredging, on habitat succession and ecological productivity in this area.1
The bottom line is that one is able to produce models that can simulate the historical,
successional patterns in these landscapes and project scenarios into the future under different management
scenarios and also different historical scenarios. And one could look at, for example, the future succession of
this area with and without the different levels of sea-level rise that may be caused by global warming. Areas
we are working on now the Putuxant River Basin in Maryland and the Everglades to try and take this
technique a little further.
The whole issue of scaling is also the focus of an EPA-funded center that is just now being
started at the Center for Environmental and Estuarine Studies called the "Multi-Scale Experimental
Ecosystem Research Center" (MEERC). The idea is to build a whole series of cosms at different scales and
to do an integrated set of experiments and models at those different scales, with the eye toward building
more sophisticated scaling models.
MEERC will look at the ways of scaling information collected at these small scales to what
we are really interested in, which is building a larger model at the landscape scale. We cannot do
experiments at that scale. Ultimately, we hope to provide better models of the real world and how these
very complex interactions between natural and human resources occur.
There are a couple of fundamental hypotheses that we are trying to test with this center.
One is that there is some relationship between the level of resolution that we study these problems at (how
finely we break up the problem both in space and in time and in the number of components that we are
looking at) and the level of predictability that is ultimately achievable from the modeling exercise or from
interpreting the data.
As you increase the resolution, the hypothesis is that the data predictability, that is the
internal pattern that you see, goes up. This is the idea behind fractals that the closer you look at a coast
line, the smaller the ruler, the larger the measured length of the coast line.
And there is a regular relationship there. But, with models, the relationship probably goes
in the other direction, that the higher the resolution, the harder it is to build predictive models.
Predictability ultimately suffers as you increase the resolution. So, modelers are always trying to reduce the
resolution, keep things simple so that the predictability is up. And experimentalists are always trying to
increase the resolution so that they can see more in the data.
And the point is, that the best models in the sense of making a trade-off between model and
data predictability for usability are ones that are somewhere in between. The bottom line on that is that
GW4-7
-------
Session GW-4
Costanza
there are severe limits. I think there will always be severe limits to the predictability or precision with which
we are able to analyze the environment.
This issue of scientific uncertainty is, therefore, a very critical one, when it comes to valuing
natural resources and also what we do with the values that we come up with for these resources. We end up
with numbers that span, maybe, an order of magnitude. Some of the things coming out now from the Chaos
Theory and other attempts to model, even what looked to be the derministic system, simply that there is a
certain level of predictability that we can expect, but it is not ever going to be 100 percent.
So what do we do with that information? We have to learn how to manage these systems
and value them. No matter how sophisticated our models are, all our models are ever going to tell us is the
range of possible outcomes. And that range, even with the best model that we can incorporate, may be very
large.
I do not think that that is necessarily a negative outcome. The question is, what do we do
with it? One of the instruments that we are exploring at the institute is the idea of a flexible environmental
assurance bond where the uncertainty is folded into the costs of a piquian kind of tax. So that the polluter
would pay not only for the known damages, but also for the potential damages, at least in a tentative way, in
the form of a bond which could be later returned if and when the damages are shown to be less than the
worst-case scenario.
This gives the proper incentives to the firms and to consumers of the products of those
firms, about the potential impacts, the possible costs, and gives them a real incentive to both reduce the
uncertainty, either by doing additional research or funding additional research, or by changing their processes
so that they are producing less damaging and less potentially damaging products.
So, in terms of ground water pollution, especially for the kinds of impacts that extend over
long periods of times and that have very high uncertainty regarding their ultimate impacts, the idea is to put
the burden of proof for those impacts on the parties that stand to gain from the impacts, not on the public.
Integrated multi-scale modeling of linked ecological economic systems is necessary to determine the worst-
case impacts. This system should provide the incentives necessary to harness market forces toward the good
of environmental protection and long-term sustainability.
REFERENCES
1. Costanza, R.H., F.H. Sklar, and M.L. White. "Modeling Coastal Landscape Dynamics,"
BioScience. Vol. 40, pp. 91-107. 1990.
2. Costanza, R.H., and L. Cornwell. "The 4P Approach to Dealing With.Scientific
Uncertainty," Environment. Vol. 34, pp. 12-20, 42. 1992.
GW4-8
-------
Session GW-4
Harmon
Liquid Assets and Paper Water; Valuation of Ground Water Under
Colorado's Prior-Appropriation System - Traditional Basis and New Issues
Eric J. Harmon, P.E.
HRS Water Consultants, Inc.
ABSTRACT
The traditional basis of ground water value under the prior-appropriation system of water
rights in Colorado is establishment of a property right to water. This requires judicial and administrative
approvals on the basis of demonstrated historic use of ground water (if tributary), ownership of overlying
land (if nontributary) as well as proof of non-injury to existing water rights. These requirements help
establish the value of a ground water right under the Colorado prior-appropriation doctrine.
Traditional ground water valuation methods are analogous to those used in real estate
appraisal. Methods used include income analysis, cost to replace a water supply, in-lieu or tap/commodity
fees, and market comparison. Each of these has advantages and disadvantages, but all are premised on the
prior-appropriation doctrine consequence of economic reward to development, consumption, and perceived
stability of supply.
New issues in ground water valuation result from many factors, including growth and
urbanization in the West, new private-sector investment strategies, elimination of subsidies, resource
depletion and contamination, technological advancement, and common-good motivations through increased
public awareness.
Future value and availability of ground water in the West will depend on consistent and
long-term policy emphasis at state and federal levels to minimize aquifer depletion, promote efficiency of
use, encourage artificial recharge, and protect ground water quality.
1.0 INTRODUCTION
The value of ground water in Colorado, as in most Western states, is tied strongly to the
prior-appropriation system within which it is administered. Establishment of a property right to ground
water requires demonstration in a court of law that water has been used beneficially. Under this system,
market value of ground water is directly proportional to its transferable consumptive use based on one or
more traditionally accepted beneficial uses. This paper examines the basis of these value concepts and
methods under the prior-appropriation system of water rights, and discusses emerging issues of ground water
valuation.
2.0 TRADITIONAL BASIS OF GROUND WATER VALUATION
2.1 Establishment of a Property Right
In Colorado, the water-rights doctrine of prior appropriation holds. This means that
appropriation, (i.e., diversion and use of water) is the first step in establishing private ownership of a right to
ground water. The earlier in history appropriation has taken place, the more "senior" the right. In times of
limited supply, the senior right has priority over junior rights to use water sufficient for its need, up to its
decreed limit. Establishing priority largely is a judicial process: one must show evidence in Colorado's
water-court system, or in some areas before a special tribunal, of a date of appropriation. In addition to
establishing a date of appropriation, one must demonstrate that a beneficial use of ground water has taken
GW4-9
-------
Session GW-4
Harmon
place as a result of appropriation. Traditional "beneficial" uses include crop irrigation, stock raising,
municipal/industrial, and commercial.1
If the use of water is heavily consumptive, that is, it results in a significant loss of water to a
watershed, the water right subsequently established may be of considerably higher value, upon transfer to a
different place or type of use, than an appropriation which is nonconsumptive. For example, flood irrigation
of row crops may consume greater than 50% of the applied water due to plant transpiration and evaporation,
whereas an industrial process using ground water may consume less than 5% of the water diverted. Much
emphasis is placed on establishing historic consumptive use of ground water in court proceedings in
Colorado. Traditionally, the average historic consumptive-use portion of the water applied is the amount
available for transfer to another place or type of use. This has resulted in a strong correlation between
historic consumptive use and value of a water right.
The process of acquiring a water right is contingent upon court approval of priority,
consumptive use, type of use (e.g., municipal), and a determination that exercise of the water right will not
injure any other previously vested water right. The adjudication process necessary to obtain a decree
conferring a water right, coupled with the prerogative of other water-right owners to lodge an objection to a
proposed new ground water right on the basis of perceived injury, often results in a time-consuming and
expensive gauntlet of application, notification, negotiation, technical and legal research and analysis, and
court trial. Time and funds expended on this process often are high: two years and several hundred
thousand dollars are not uncommon. For example, a recent large proposal to pump ground water and export
a portion from its basin of origin was applied for in late 1986, came before the court for trial in late 1991,
and entailed expenditures (by both applicant and objectors) of several million dollars. The court decided
against the applicant in that case, which currently is being appealed. If the outcome is favorable in such a
proceeding, a decree for a water right is conferred on the applicant.
A water right is a property right1 analogous to real estate or, perhaps closer to the point, a
mineral property. A conditional decree (which can be conferred before one has placed all water claimed to
a beneficial use) roughly is equivalent to a mining claim; an absolute decree (when all of the water claimed is
in use, and so noted by the court) is analogous to a mineral patent.
A decree, whether conditional or absolute, often significantly enhances the value of water,
though most water-rights decrees do not specify consumptive use or address very real concerns about
administrative or physical constraints to use of water. It is thus possible (and not uncommon) to encounter a
sheaf of court decrees which, at face value, confers upon the owner title to thousands of acre-feet of water,
while reality is that low to non-existent historical consumptive use, or physical and administrative constraints
to that use, make the water rights all but worthless. In Colorado this type of "paper water" has a long and
distinguished history of flimflam and chicanery. It takes considerable engineering and legal work to sort out
water-rights fact from fantasy.
22 Physical and Administrative Constraints to Value
Historic consumptive use is, as discussed above, a primary determining factor in the value of
a ground water right. A related factor is transferability of water. For example, a water right historically
used for irrigation has established a pattern and volume of use over many years. Since not all of the water is
consumed by irrigation, there is a significant volume of irrigation water that returns to the watershed as a
return flow. Junior-priority decreed water rights very likely have been established which depend, either
wholly or in part, upon the presence and regularity of these irrigation return flows. Transfer of a water right
to a different place of use or type of use then may have an injurious effect on a junior right. Transferability,
therefore, typically is limited to the historic consumptive use of the right. Some watersheds in Colorado
under jurisdiction of locally controlled boards have adopted rules which prevent or severely restrict such out-
of-basin ground water transfer due to the possibility (real or imagined) of injury to in-basin rights. Such
actions typically reduce regional market value of the ground water, but may in turn enhance value for in-
basin use.
GW4-10
-------
Session GW-4
Harmon
A special type of ground water in Colorado is that which has been determined, either by
administrative rulemaking or by judicial notice, as "nontributary". Under Senate Bill 5, enacted by the
Colorado legislature in 1985, nontributary ground water "means that ground water . . . the withdrawal of
which will not, within one hundred years, deplete the flow of a natural stream ... at an annual rate greater
than one-tenth of 1% of the annual rate of withdrawal" (37-90-103 (10.5), Colorado Revised Statutes).
Entitlement to such water is based on overlying land ownership, thickness of water-bearing aquifer material,
and the propensity of the material to yield water. Since nontributary water is administered independent of
the priority system of water rights which governs surface water and shallow (tributary) ground water, no
junior rights have been established which are dependent on a percentage of consumptive use and return flow.
Thus nontributary rights are usable and reusable to the full volume of the water right, a factor which
significantly enhances their market value and encourages full withdrawal of the water-right owner's
entitlement.
Paradoxically, the fully-consumptive/fully-reusable nature of nontributary water enhances
value of a largely nonrenewable resource. This factor, coupled with a statewide policy of 100-year nominal
aquifer life (a water user is allowed to pump 1% of the computed available aquifer volume each year)
promotes withdrawals of ground water far beyond the volume of water reaching the aquifers each year by
natural recharge. Water levels in deep nontributary aquifers in the Denver Metropolitan area, for example,
are declining since pumping exceeds recharge. Rectification of policy encouraging "mining" of ground water
is one legislative and administrative challenge which faces the State. One possible solution would be to
encourage or even require artificial recharge of ground water, though technology in this area currently lags
far behind the need. Another would be to curtail pumping until a recharge/discharge balance is achieved,
though this would be virtually impossible to effectuate without a significant increase in water conservation
and/or renewable supplies to replace current dependency on nonrenewable, nontributary ground water.
As might be imagined, chemical quality of ground water is a primary determining factor hi
value of ground water. Though the vast majority of economically feasible ground water hi Colorado is of
good quality, there are aquifers and locales where there are valid concerns about ground water quality. Most
dissolved constituents found in Colorado ground water are naturally occurring; many are benign but
undesirable. For example, relatively low levels of dissolved iron and manganese are nontoxic but are
aesthetically displeasing, causing orange or brown water and stained fixtures. Though most undesirable
chemical constituents hi ground water are treatable by a variety of methods, costs of infrastructure and
treatment processes can be high enough to cause a decision to abandon development of an otherwise-suitable
supply of ground water. In an income approach to valuation a future stream of disbursements for treatment
may be high, necessitating a high sale or lease price to sustain a desired rate of return on a water-rights
investment.
Proximity to water market is another determining factor in value of a ground water right.
Transportation costs for water, particularly in interbasin or intermountain transfers, can be high due to high-
relief Colorado topography, large distances, and difficulty of obtaining necessary judicial and administrative
approvals. If an out-of-basin ground water transfer once is approved and constructed, however, the water
right value is enhanced. It is considered "new" water hi the recipient watershed, and is therefore fully
reusable and 100% consumable. Economies of scale prevail hi water transport: unit transport costs are
much lower for large-volume projects than for small-volume ones.
23 Ground Water Valuation Methodologies
Traditional methods of ground water valuation hi the priority-system dominated Colorado
water-rights market by and large have followed standard methods for real property appraisal. Commonly
used methods of ground water valuation include income-stream analysis, cost to replace a water supply,
tap/commodity fees or ui-lieu fees, and market-transaction comparison.
An income approach, also termed analysis of net present value, has been used hi valuing
ground water where development plans exist which are specific enough that an income stream can be
estimated over a reasonable project lifetime (typically 40 years). There are several difficulties with this
GW4-11
-------
Session GW-4
Harmon
method, however. Predevelopment plans often are in an embryonic state at time of need for appraisal of
ground water value. Often, too, it is difficult to predict with accuracy future demand for a water resource
based on population growth and business climate. This leads to large uncertainties in potential income
resulting from ground water ownership.
Perhaps the most difficult task in performing an income analysis of ground water value is
establishing a reasonable discount interest rate. Discount rates reflective of undeveloped ground water must
reflect risk in legal status of water, physical constraints to development, and uncertainty of a sale talcing
place. Too often sensitivity of ground water value to an appropriate discount rate is not fully appreciated.
With insufficient emphasis placed on proper determination of rate, the result may be an erroneous valuation.
Or, worse, an unscrupulous appraiser conceivably could select a discount rate to force outcome to a
predetermined value. Part of the problem is that water-rights transaction data typically is considered
proprietary, so that little comparative rate information exists.
Selection of a discount rate for water-rights valuation should be based on an informed
estimate of risk and an informed comparison with alternative investments of comparable risk.
Another method used to value ground water is analysis of cost to replace a ground water
supply with an alternative water supply. Difficulties arise when water-supply alternatives, as is often the case
in the arid West, are unsuitable or virtually unavailable.
In-lieu fees or tap/commodity fees have been used as a proxy for valuation of ground water
supplies. In-lieu fees are fees charged by municipalities to private developers who either do not have, or who
elect to retain, ground water rights associated with land subject to annexation. Tap and commodity fees are
fees charged to water users upon hookup and use of a municipal or industrial supply. These comparisons,
though relatively well accepted in the water community, may be inaccurate because fees often are set
artificially high as disincentives to private sale of water rights apart from annexed land. Further, fees can be
set as disincentives to development and thus are not reflective of open-market value for ground water.
Market comparison is perhaps the most common and most readily accepted traditional
method of ground water valuation in the Colorado system of water rights. The open-market database of
water-rights transactions is limited in size and is anecdotal: there is no common, public database of
transaction data. Differences among ground water rights include location, historic consumptive use, decree
limitations on amount or type of use, administrative restrictions, and physical constraints.
Differences most often are apparent only upon detailed study of transactions and an in-
depth understanding of water rights. Moreover, many water-rights transactions are not "arm's length" sales
or leases. It is thus difficult to establish comparability. Notwithstanding these potential problems, however,
market comparison remains the best-accepted method of ground water valuation in condemnation
proceedings in courts of law.
The outcome of traditional methods of analysis is a result of the prior-appropriation
doctrine: value is enhanced by development, use, and more or less continuous consumption of ground water
over long periods of time.2 Only recently have new or nontraditional issues been considered in valuation of
water rights. Enhancement or preservation of wildlife and/or wetland values has, encouragingly, begun to
play a part in determination of highest and best use of ground water resources. Likewise, conjunctive use of
renewable and nonrenewable water, and preservation of community values (e.g., revegetation of farmland
from which water was diverted for sale) are beginning to be considered in valuation of ground water. This
should not be interpreted as an indictment of the doctrine of prior appropriation: aspects of this system
encourage efficient and nonwasteful use of the resource. However, it will be necessary to consider these and
other non-traditional factors hi ground water valuation if we are to continue to enjoy use of ground water
without severe and irreparable degradation and diminution of the available supply.
GW4-12
-------
Session GW-4
Harmon
3.0 NEW ISSUES IN GROUND WATER VALUATION
3.1 Emerging Ground Water Investment Strategies
New strategies for investment in water rights include syndication. A private trust is created
to invest capital in water rights with the single objective of making a profit.3 This tends to promote
development of a detailed understanding of water rights to be included in a portfolio, but at the same time
tends to promote speculation. Colorado water law now discourages speculation, requiring evidence of a
contract or some other form of written agreement for an actual use of water prior to conferring a decree.
Syndication also has led, in recent years, to an increase in foreign investment in water rights.
Reduction or elimination of government subsidies also has had an effect on water rights
values. With virtual elimination of large subsidized water projects, particularly dams and reservoirs, more
attention has been focused on ground water resources. Water-development projects have experienced an
increase in private funding at the same time as an increase public awareness regarding water issues and
demands for environmental safeguards. These factors, coupled with increasing water demand in Sunbelt
states, have made unit costs of new water supplies much higher than continuing reliance on existing sources
of supply. This situation has led to a relatively continuous rise in water-rights values in the West.
3.2 Resource-Effect Considerations
Assessment of damage to ground water resources has been slow to develop because of
several factors: (1) lack of public concern about ground water resources, (2) high cost of data collection and
analysis, (3) incomplete scientific understanding of severity of damage, and (4) inconsistency and misdirection
of government policy. Damage may include ground water contamination, spinoff effects such as wetlands
degradation, as well as long-term drawdown effects due to prolonged ground water withdrawal. Developing
a detailed understanding of aquifer systems is complex and costly, and typically has been undertaken only
when severe contamination already has taken place. Analyses of this type have been site-specific in nature,
and have not addressed larger ground water systems within which affected areas lie. An effort to assess
ground water systems on a large scale has begun with the U.S. Geological Surveys series of studies
culminating in the RASA (Regional Aquifer System Analysis) group of publications.4 This series is not yet
complete but represents a beginning toward a working understanding of resource-damage effects on ground
water systems.
Mitigation requirements affect ground water value. There is a perception of decreased
short-term value to ground water if mitigation of a ground water damage problem is necessary: certainly this
is not encouraging to private investment. Also discouraging to private investment in ground water is
increased long-term administrative burden, uncertain administrative authority, and difficulty in establishing
agreeable terms of stewardship or mitigation. Value of ground water thus declines, often drastically, if a
damage problem requiring mitigation is perceived.
Due to uncertainties in damage assessment and mitigation, estimating future costs of a
ground water investment venture is difficult. Some entities, particularly public water providers, have
attempted to value ground water within a framework of mitigation and development tradeoffs by contingent
valuation. Polling of water users is done in which respondents are asked to place a dollar value on individual
willingness to pay, for example, costs of large-scale artificial ground water recharge to enable ground water
development to avoid dam-building in a pristine area.
Contingent valuation, though useful in assessing public perceptions and values concerning
ground water, may risk slanted valuation results. This is particularly true if emotional, "hot-button" issues are
the basis for willingness-to-pay comparisons, or if the public is not particularly well informed about potential
risks inherent in the alternatives.
GW4-13
-------
Session GW-4
Harmon
33 Time-Of-Need or "Trigger" Arrangements
Time-of-need arrangements, particularly drought-year leasing, are receiving much attention
in Colorado and throughout the arid West. A typical scenario involves a lease contract between a municipal
water provider and an irrigation water user. The irrigationist pursues his/her usual pattern of seasonal water
diversions for crop irrigation during wet- or normal-precipitation years. During a drought year irrigated land
is left fallow (or is planted in crops which are nonconsumptive of water), and water instead is diverted for
use by the municipality. Alternatively, the irrigationist may forego his normal surface-water irrigation supply
during a drought year and use ground water instead. Typically an annual retainer is paid to the irrigationist,
along with a premium during drought years when the lease arrangement is exercised.
A variation of this drought/trigger arrangement involves pumping a non-renewable ground
water source only a few days or weeks during peak-demand season. This enables a relatively low-priority
renewable-source water right to be exercised for municipal or M&I use. Both of these trigger arrangements
tend to enhance the value of a ground water right. At the same time it encourage conservative use of the
ground water resource.
Economies of scale prevail in such arrangements. The cost of administrative and judicial
approvals for water transfers may be high to show proof to other water users that no injury will result. Also,
if physical facilities are necessary for water diversion or conveyance, large-scale projects tend to be more
economical in terms of cost per unit of water delivered.
3.4 Technological Advancement
Ground water use, as with most resource development and allocation, increasingly depends
upon application of new technology. These advances primarily will take the form of improved allocation,
conservation, and efficiency of use, with the goals of ground water supply longevity and protection of quality.
Less emphasis will be placed on improving wells, pumps, and the like, simply because improvements to these
mechanical constructs will be measured in terms of fractions of 1% of marginal efficiency increase.
Currently there is much more room for improvement in allocation, protection, and conservation of ground
water than there is for better wells or pumping equipment.
Such changes, while perhaps perceived by investors in the short term as diminishing ground
water value, will, over a longer term, tend to enhance values by promoting aquifer longevity and preservation
of water quality.
Artificial recharge of ground water is one new technology which must experience a
tremendous increase in both public and private funding for research and development if "mining" of
nonrenewable or slowly-renewable aquifers is to be alleviated. Many areas of the West are experiencing
rapid decline in water levels as a result of ground water pumping far in excess of natural recharge. Many
western states, where authority for administration of ground water traditionally resides, have policies of finite
aquifer life. Planned "lifetimes" of aquifers have ranged from as little as 20 years to as much as 300 to 500
years, though pumping costs, drawdown interference between wells, and supplemental-well costs may
effectively limit an aquifer's economic life to a fraction of these nominal terms.6
There are several problems traditionally associated with ground water recharge to deep
aquifers by systems of wells. One is relatively high pumping cost to reinject water into a deep aquifer.
Another is the tendency of naturally occurring dissolved chemical constituents in injected ground water to
precipitate out of solution upon reinjection, causing well and formation plugging. A third problem is the
tendency of some aquifers to irrecoverably compress upon initial withdrawal of ground water, limiting the
volume of pore space available for storage of newly recharged water. Each of these problems may be
solvable (to varying degrees) if sufficient attention is paid these issues.
In a positive sense, the potential for ground water recharge to be used in lieu of surface-
water storage reservoirs is extensive. Ground water storage "reservoirs" tend to be environmentally much
GW4-14
-------
Session GW-4
Harmon
less damaging and far less expensive than surface storage. Evaporation losses, which are high in the arid
West, effectively are zero in a ground water storage reservoir.7 Artificial recharge of ground water also
tends to enhance yield and value of surface-water rights, and can promote preservation of wetland areas by
balancing seasonal water supply and demand.
3.5 Motivations for Allocation and Protection
Consistent and well-designed policy concerning ground water, on both state and federal
levels, must be driven by public concerns about availability, cost, and quality of water supplies. In a larger
sense, threats to livelihoods and community values traditionally must be perceived and communicated to
cause effective change. Public education concerning ground water, therefore, is key in developing coherent
policy. Because a majority percentage of ground water in western states is consumed by irrigated agriculture,
incentives should focus on promoting efficiency, flexibility, conservation, reuse, and ground water recharge in
this sector.8 In Colorado, economic considerations, user cooperation, and administrative requirements are
yielding progress toward these goals.
Increasing public awareness of the shortsightedness of ground water policies based on
mining nonrenewable or slowly-renewable supplies should be a top priority. Such policies contradict the idea
of indefinitely-sustainable growth and development in the West. Several key western cities, notably
Albuquerque, Tucson, and Denver to a lesser degree, already have experienced economic effects of water-
level declines due to excessive pumping. In most aquifers, however, the skies (and water levels) are not
falling rapidly, and public awareness should be coupled with education as to the value of ground water if it is
used in an intelligent and well-planned manner.
Currently there is increasing public awareness of community impacts to shortsighted water
policy. Factors receiving attention include fears of loss of livelihood, loss of arable land, and centralization of
water ownership or control. Public awareness also has focused, recently, on concepts of reserved water rights
(particularly for headwaters wilderness areas) and instream flow rights. With regard to ground water,
emphasis must also be given to public awareness of sustainable levels of ground water recharge (both natural
and artificial), use of ground water and surface water interactively, and the fact that rapidly declining artesian
water levels are, in many cases, not indicative of a significant loss of water in aquifer storage.
Increasingly, public perceptions of ground water and surface water in the West have shifted
toward water as a "common" rather than a commodity subject only to individual ownership.9 Strict prior-
appropriation administration of water rights in the West thus appears to be shifting toward concerns for the
common good. An example of this is newly enacted legislation in Colorado requiring revegetation of
formerly irrigated lands taken out of production upon transfer of a water right to a different point of use.
Another example is a relatively new requirement, in Colorado, that a water user must provide evidence of a
sale or lease for an actual use prior to conferral of a water-right decree, as opposed to honoring private
speculation as a legitimate use for water.
4.0 CONCLUSIONS
There is no single, traditional valuation method which is universally applicable for appraisal
of ground water within prior-appropriation systems of water rights. Considering only the traditional basis for
comparison of ground water values, data are too sparse and varied, and physical situations too diverse, for a
single method always to be applicable. A detailed knowledge of the physical setting of ground water, history
of use, and water rights concerning the supply, are necessary to form a full and reasoned understanding of
ground water value.
Future ground water value and availability in the West will depend on adoption of forward-
looking attitudes and policies. These must include consideration of ground water artificial recharge,
cessation of rapid mining of nonrenewable supplies, promotion of efficiency, conservation, and reuse of
water, as well as aquifer damage assessment and mitigation.
GW4-15
-------
Session GW-4
Harmon
Preservation of long-term value in water rights probably will require redefinition of the
concept of value to account for shifts in water-supply availability and quality, changing public perceptions,
and must, at all costs, consider recharge-based ground water withdrawal limitations.
5.0 REFERENCES
1. Vranesh, G. Colorado Water Law. Design Press, Boulder, CO. pp. 120 and 141. 1987.
2. Coggins, G.C. and C.F. Wilkinson. Federal Public Land and Resources Law. University
Casebook Series. The Foundation Press, Mineola, New York. pp. 359 and 371. 1987.
3. Harvey, E.F, "The Emerging Applications of Water Valuation," Colorado Water
Engineering and Management Conference Proceedings. Colorado State University, Fort
Collins, CO. pp. 418-422. 1987.
4. Sun, R.J. Regional Aquifer-System Analysis Program of the U.S. Geological Survey.
Summary of Projects. 1978-84. 264 p. 1986.
5. Otto, J. "Reclamation Bureau Takes Welcome Turn In Policy," Editorial in The Rocky
Mountain News. December 27, 1988.
6. Vaux, H.J. "Economic Aspects of Ground Water Recharge," Artificial Recharge of Ground
Water. Butterworth Publishers, Boston, MA. pp. 703-718. 1985.
7. Asano, T. "Overview: Artificial Recharge of Ground Water," Artificial Recharge of
Ground Water. Butterworth Publishers, Boston, MA. p. 5. 1985.
8. MacDonnell. Integrating Tributary Ground Water Development Into The Prior
Appropriation System: The South Platte Experience. Natural Resources Law Center,
University of Colorado, Boulder, CO. pp. 42-46. 1988.
9. Murray, M. "Colorado's Venerable Water Law System Isn't Meeting Its Modern
Challenges," Editorial in The Denver Post. November 25, 1989.
GW4-16
-------
Session GW-4
Crutchfield
Issues in Measurement of the Non-Use Value of Ground Water
Stephen R. Crutchfield3
Economic Research Service
U.S. Department of Agriculture
ABSTRACT
Considerable public attention has focused lately on the need to protect the quality of ground water resources.
The considerable public and private resource commitment to protecting ground water quality is evidence of
the value society seems to place on this resource. One component of the total social value placed by society
on clean ground water may be the values placed on the resource by non-users. This paper reviews the
economic issues surrounding the measurement of non-use values of natural resources, including option,
existence, and bequest values. These concepts are further explored in the context of valuing protection of
ground water from chemical contamination. The factors that complicate the measurement of use value of
ground water resources, particularly uncertainty about pollutant transport, environmental fates, and eventual
human health effects also complicate efforts to specify and quantify non-use values. The paper concludes
with some possible approaches to measuring non-use value in practice.
1.0 INTRODUCTION AND OVERVIEW
In the past 10 years, a considerable amount of public interest has arisen about the quality of
the nation's ground water resources. This is especially true for agricultural chemical residuals which may
potentially degrade ground water quality. Discovery of nitrates and pesticides in ground water during the
late 1970's and early 1980's dispelled a commonly held view that ground water was protected from these
chemicals by layers of rock, soil, and clay. Recent estimates indicate that at least some quantities of 46
pesticides have been detected hi drinking water wells hi 25 states.25
Ground water is an important source of drinking water, especially in rural areas. Concern
about ground water contamination, at least from the agricultural policy area, is driven by fears that exposure
to agricultural chemicals in drinking water may pose a risk to human health. In response, substantial
resources, both federal and state, have been devoted to protecting ground water from chemical
contamination. We have seen billions of dollars targeted to cleaning up hazardous waste sites under the
Superfund program. Under the President's Water Quality Initiative the Agriculture Department, EPA, and
NOAA treat the protection of ground water from agricultural chemicals as a primary objective.
Our understanding of the economic benefits of protecting ground water quality and the costs
of policies to prevent ground water pollution are sketchy at best. Relatively few studies which specifically
estimate the value consumers placed on reduced risk of chemical contamination hi their drinking water wells
have been done. (See Doyle, et. al.,8 Edwards,9 Shultze and Lindsay,19 and Sun.22) Other studies take a
cost-avoidance or averting expenditures approach to valuation, rather than any estimation of willingness to
pay for ground water protection (the best known example is Nielsen and Lee ).
Despite all the interest in ground water pollution, there is some evidence that, for some
classes of pollutants, severe ground water quality impairments may not be all that widespread. The
Environmental Protection Agency recently completed a nationwide survey of drinking water wells which
aStephen Crutchfield is the Leader of the Environmental Valuation Section, Resources and Technology
Division, USDA Economic Research Service. The views expressed in this paper are those of Dr. Crutchfield,
and do not necessarily reflect official USDA policy. The contributions of Joseph Cooper and Daniel
Hellertstein of ERS in drafting the survey question in Appendix I are gratefully acknowledged.
GW4-17
-------
Session GW-4
Crutchfield
showed that 1.2% of community water systems and 2.4% of rural private domestic wells sampled in the
survey contained nitrates at levels higher than the EPA's recommended levels. About 10% of community
wells and 4% of domestic wells surveyed contained detectable levels of one or more pesticides, but the
concentration levels were, for the most part, below levels considered by the EPA to pose a risk to human
health. Could it be the case, though, that non-use values of protecting ground water quality may exist?
Given the considerable uncertainty surrounding the scope and extent of actual risks to human health from
chemical contamination of water supplies, it may be that the public concern is indicative of value placed on
protecting ground water quality by non-users, or reflects a desire to protect future uses of ground water by
preventing any further degradation.
This paper takes a brief look at the issue of non-use values as they pertain to ground water
quality. The objective is to look at economic valuation methods when non-use values are thought to exist for
a natural resource, and relate these to the specific case of ground water resources. The characteristics which
differentiate ground water protection from other natural resource preservation issues are also discussed.
Given the complexity of the concepts and the substantial literature on non-use valuation, this paper cannot
present an exhaustive review of the entire issue. Rather, it is intended to form the starting point for a
general discussion of whether non-use values for ground water protection are important, and suggest some
possible ways to place some economic measures on these values.
The outline of this paper is as follows. Section 2.0 briefly reviews some commonly held
economic measures of non-use values. Section 3.0 discusses how these measures can supply useful
information to analyze policies to protect natural resources. Section 4.0 relates these measures to the
particular resource issue in question, and shows how the characteristics of the ground water quality problem
complicate the valuation task. Section 5.0 concludes the paper, and offers some suggestions for further
research.
2.0 CONCEPTS AND MEASURES OF NON-USE VALUE
Economists have long recognized that natural resources and environmental amenities may
be valued even by those not directly using these resources. Weisbrod and Krutilla are seminal articles in
this regard. The taxonomy of non-use values varies among economists, but generally these values are placed
in two categories: Option Value, and Existence Value. Within each category, subclassifications can be
found. (A complete technical development of all the non-use value measures is, of course, beyond the scope
of this paper. My intention here is to give an overview of the main issues in simple, descriptive terms. The
interested reader is referred to cited literature for more detail.)
2.1 Option Value
Weisbrod, in his discussion of provision of public goods such as national parks, noted that in
addition to current users of a park potential users who were not current users could still place a value on
retaining the option to use a park at some future date. He called a willingness to pay to preserve the
opportunity to use the park in the future option demand.
Option value is a measure of the benefits of a change in the price, quantity, or quality of a
resource under uncertainty. Option value is related to option demand, and is derived in the following
manner.3>5'15 First, calculate the compensating variation of a price change in each state of the world, and
take the expected value (ECV). Second, find the fixed payment (option price, OP) that will leave an
individual indifferent in terms of expected utility between the status quo and receiving the change but making
the payment. Option value, then, is the difference between the two: OV = OP - ECV.
Option value is sometimes divided into two sub-topics: demand-side option value, and
supply-side option value.
GW4-18
-------
Session GW-4
Crutchfield
22 Demand-Side Option Value
The term demand side option value refers to a situation where there is uncertainty over
some argument to an individual's indirect utility function, and the objective is to value the benefits or costs of
a project that changes the some other argument in the indirect utility function. Formally, if we denote the
probability distribution of the uncertain parameter (such as a substitute good) as -K , and En- is the
expectations operator over the probability distribution TT , then we can define demand side option value in the
following manner:3
OVD = OP-Br*CV
23 Supply-Side Option Value
On the other hand, the uncertainty may be in the quantity or quality of a good in the future.
Suppose we let 0 represent the base state of the world, and 1 represent a new state where some action has
been taken which affects resource availability. If n° represents the probability distribution of the uncertain
argument to the individual's utility function in the base case, and JT 1 represents the distribution post-change,
the value is measured as the difference between option price OP and the change in expected consumer
surplus:
OVS = OP-E(7r°-7r1)*CV
2.4 Existence Value
Basically, existence value is the value placed by an individual on the fact that a particular
resource or environmental condition exists, regardless of whether or not that person obtains any direct
services or use from that resource. Many different types of existence value have been proposed, including
(but not limited to):1'2'16'17'20
• Vicarious consumption value: an individual derives value from knowing that others
may consume or use the resource
• Bequest value: The value placed on preserving natural resources/environmental
amenities for the future: "pass it on to the grandchildren"
• Intrinsic value: an individual values a resource simply because it exists.
3.0 WHICH NON-USE VALUE DO WE USE: EMPIRICAL ESTIMATION ISSUES
3.1 Option Value vs. Existence Value
Economists have debated the concept of option value for many years. Much of the debate
has centered on whether or not OV is positive or negative, and what determines the sign of OV. Some
evidence suggests that OV may be a large component of total resource value. Option price, which is a
determinant of option value, has been used as a measure of resource value in the context of ground water
quality8'9'19 and surface water quality.7 When considering non-use values, however, there has been some
confusion in the past over whether option value is some additional component that should be added on to
use value. Smith '21 and Freeman, make clear that option value should not be combined with use values.
On the other hand, Callaway notes that supply side OV becomes an appropriate measure of non-use values
when the issue is uncertainty in resource quality, which is an issue we are concerned with here: "Supply-side
option value . . . becomes the appropriate metric to measure changes in non-use values when the effects of
changes in air pollution on a resource are characterized by a distribution of consequences."4
With regard to existence or bequest values, one problem is their obvious nature: "Unlike
option value, existence or bequest values are more empirical in nature: in most cases, we merely assert that
GW4-19
-------
Session GW-4
Crutchfield
they exist and move on to the question of their measurement." Much of the discussion in the literature has
revolved around the nature of the existence value being estimated: whether it is pure existence value,
indirect forms of use, vicarious consumption, and so forth. Much has been also made of the role of motives
in determining non-use value, particularly by Brookshire et. al. who assert that with the exception of intrinsic
value all other motives for existence value reduce to a form of option value or are inconsistent with the
economic efficiency paradigm which underlies benefit-cost analysis.
3.2 Guidelines for Defensible Valuation
Much of the work on existence value has been done in the context of recreation and wildlife
valuation. A fundamental problem here is that the assumption of weak complementarity does not hold for
the case of non-use or existence values, and so value cannot be inferred from the demand for complementary
goods. This implies that "non-use and/or existence values cannot be determined from information about the
behavior of users that is obtained from travel cost models; the only way true non-use values can be estimated
for users and non-users alike is through contingent valuation methods."3
The use of CVM to value ground water has been discussed earlier in this conference, so a
complete discussion of its applicability will not be explored here. However, some issues can be raised here.
Freeman,13 following on work by Cummings, et. al.6 has established what he calls "Reference Operating
Conditions" for designing a reliable CV instrument for estimating non-use values. Three in particular bear
mentioning:
1. The instrument should clearly and accurately describe the change in the quality or
availability of the resource being valued.
2. The change in quality or availability being valued should be in the range of
experience of the respondents.
3. The instrument should avoid questions that are framed in such a way as to link the
survey instrument to minimize the likelihood of strategic behavior, zero protest bids,
and nonrespondents.
These conditions may prove troublesome if we apply this to the particular case of ground water (a matter to
be addressed next).
4.0 GROUND WATER AS A SPECIAL CASE
4.1 Nature of the Resource
Much of the non-use value work has been done in the context of environmental "goods",
such as the availability of clean water for recreation, preservation of endangered species, preservation of
wilderness, and so forth. When the issue is ground water quality, however, the "good" in question has
different characteristics. The context within which we are operating is primarily the disamenity associated
with human exposure to hazardous substances in their water supplies. There is a higher level of abstraction
and unfamiliarity involved in determining people's attitudes towards potentially hazardous chemicals in an
unseen resource; contrast this with asking people in CVM context how they feel about tangibles such as
grizzly bears, goose hunting, or maintaining wild environments. What we are actually valuing here is not so
much the resource itself (water) but people's appraisals of the risk of exposure to hazardous substances.
Interestingly, in its policy approaches to protecting ground water quality, the EPA has
proposed a classification scheme which fits into our discussion of option value and existence value. In its
classification scheme, the EPA has defined "Class I" ground water resources as those having some unique or
irreplaceable ecological value, and are afforded the highest level of protection. Class II resources are those
currently or potentially used for human and/or animal consumption (italics mine), and are afforded relatively
less protection. Class III ground water supplies, those already irretrievably contaminated, are afforded the
GW4-20
-------
Session GW-4
Crutchfield
least protection from contamination. At first glance, it would seem that existence values could be important
for Class I resources, and that option price analysis could be applied to situations where the issue is
protection of potential water supplies.
4.2 The Importance of Uncertainty and Irreversibilitv
Despite the considerable interest in assessing the value to users and non-users of preventing
ground water pollution, applying non-market valuation techniques to this issue is complicated by the nature
of the resource. Three characteristics of ground water contamination in particular stand out:
• There is considerable uncertainty about the linkages between human action (farm
chemical use, for example) and environmental endpoints (eventual contamination of
an aquifer.) This uncertainty has a spatial dimension: pollutants may flow in
plumes through an aquifer, and well contamination may evidence far from the point
of chemical spill. There is also a temporal dimension: it may take a considerable
amount of time for contaminants released at the surface to percolate through the
soil to reach the aquifer.
This is relevant in the context of option price or option value measurement. It may
be difficult to define and assign probabilities to the alternative states of the world
and resource conditions. The sequence leading from policy choice (banning
leachable farm chemicals, for instance) to change in resource condition
(contamination levels in the aquifer) is attenuated by the uncertainty surrounding
the fate and transport of pollutants in the soil and underlying media. Since the
good in question is not really the amount of water supplied but the amount of
chemical present in the water at some future point in time, it may be difficult (in
terms of the notation used above) to pin down the n1 distributions. Thus, it may be
difficult to specify the change in the quality or availability of the resource being
valued in a CVM question in a manner that is clearly linked to policy choices made
ex-ante. This would seem to run afoul of Freeman's condition #1, above.
• The linkage between objective measures of ground water quality and resulting
usability of the resource by humans is also ill-defined. There are clearly instances
where hazardous substances have made ground water resources unsuitable for
human consumption. However, particularly for the issue of agricultural chemicals
in ground water we are faced with the task of placing values on exposure to
substances of unknown risk potential at low concentrations. The EPA has
established maximum contaminant levels for many chemicals in drinking water, but
our understanding about a cause-and-effect relationship between exposure and
health effects is still cloudy, especially for low level exposure over long periods of
time. Again, this raises problems for condition #1, in that the change in quality of
the resource (which for altruistic or nonvicarious consumption values are likely to
be driven by changes in health risk) may not really be readily describable.
• Ground water may be irreversibly contaminated when chemical concentrations
begin to rise. The received wisdom seems to be that, at least for some substances,
remediation or cleanup is either too difficult, or too costly compared with
alternatives such as provision of bottled water, to be practicable. Estimation of
option price or option value might be a possibility here: "Option value now is
considered to indicate the importance of the selection of an ex ante versus an ex
post perspective in the definition of benefits resulting from a change in some
dimension of environmental quality under uncertainty."7
GW4-21
-------
Session GW-4
Crutchfield
43 The Role of Health Effects
As stated earlier, much of the concern about ground water contamination is based on the
concern that human exposure to chemicals in drinking water may lead to health risks. One component of
non-use/existence value which seems to apply in this case is an altruistic or vicarious consumption value
associated with preventing ground water contamination: individuals may not want friends, family members,
or others exposed to potentially hazardous substances which could harm their health. It would seem that
this sort of non-use value would be the most reasonable type of valuation we could consider.
However, we run into the difficulty of precisely quantifying and valuing these health risks,
and expressing them in a way that is comprehensible to subjects in a CVM study, for example. It's one thing
to show a person a picture of a grizzly bear; its quite another to talk about marginal increments in cancer
risk from aldicarb concentrations in excess of 5 parts per billion.
Communicating health risks in an understandable manner is a problem for many resource
valuation questions. My own personal experience from reading press releases from environmental groups
and attending briefings before Congressional subcommittees is that the scientific uncertainty and imprecision
about health effects tends to get reduced to sound-bite simplicity when presented to the lay public. We
either see such statements as "we don't want cancer-causing chemicals in our water at all" on the one hand,
or "farm chemicals are safe and effective when used properly and do not pose a risk to your health" on the
other. Our ability to frame questions of sufficient precision to elicit from respondents the value of marginal
changes in the concentration of potentially hazardous substances present in unknown amounts in unseen
water resources is rather limited. Our staff at ERS is currently wrestling with the wording of CVM
questionnaires designed to value the extra margin of safety associated with reducing chemical concentrations
in drinking water.
Assessing the use value of reducing or preventing these risks is daunting enough: the
hazards and pitfalls of valuing health risks of uncertain or imprecise magnitude are well known. Trying to
assign values to non-use, altruistic, or vicarious consumption values to reductions in health risks which accrue
to others only compounds the problem. These problems of uncertainty in risk communication raise concern
whether a CVM question designed to elicit non-use values could satisfy for Freeman's Reference Operating
Conditions #2 and 3, above.
5.0 CONCLUSIONS AND SUGGESTIONS FOR FUTURE RESEARCH
It is conceivable that non-use values of clean or protected ground water, particularly
altruistic and vicarious consumption values, could be discerned through a well-crafted empirical study. The
Economic Research Service has already begun a modest effort to survey residents in rural areas to determine
the respondents' valuation of some rather generic levels of ground water quality. This survey is not, of
course, aimed at directly estimating non-use values per se, but responses to the survey, when broken down by
user/non-user classification, could shed some insight in this area.
One approach we are considering in designing our survey instrument is to make the
reference points which define water quality generic rather than specific. For instance, we define three levels
of water quality: no contamination, "safe" (i.e., concentrations of chemicals below recommended levels) and
"other" (i.e., concentrations above "safe" levels). The aim of the questionnaire is to elicit people's willingness
to pay for additional margins of safety, without having to precisely define just what "safe" levels are.
However, I would argue that one should proceed extremely cautiously before attempting to
attach too much weight to measuring non-use values. Five years ago, Smith concluded "Given the current
state of knowledge, prudence suggests that appeals to substantial nonuse values in benefits analysis for
environmental resources (that are not unique) may not be warranted . . . Before the relationship between
measures of use and non-use values can be established, it will be necessary to define how individuals perceive
the specific terms of availability of the resources involved as well as how these perceptions are influenced by
uncertainty."20 Even though we have had considerable success in recent years developing sophisticated non-
GW4-22
-------
Session GW-4
Crutchfield
market valuation methodologies, the difficulties posed by this particular resource issue makes Smith's
cautionary note equally relevant today.
6.0 REFERENCES
1. Brookshire, D. and V.K. Smith. "Measuring Recreation Benefits: Conceptual and
Empirical Issues," Water Resources Research. Vol. 23, No. 5, pp. 931-935. May 1987.
2. Brookshire, D., L. Eubanks, and C. Sorg. "Existence Values and Normative Economics:
Implications for Valuing Water Resources," Water Resources Research. Vol. 22, No. 11,
pp. 1509-1518. October 1986.
3. Brown, G. and M. Plummer. "Non-Market Measures of Non-Use Values," Acidic
Deposition: State of Science and Technology - Volume IV: Control Technologies. Future
Emissions, and Effects Valuation. The US National Acid Precipitation Assessment
Program. Washington, pp. 27-75 to 27-82. 1991.
4. Callaway, J. "Estimation of Non-Use Values for Acidic Deposition/Air Pollution," Acidic
Deposition: State of Science and Technology - Volume IV: Control Technologies. Future
Emissions, and Effects Valuation. The US National Acid Precipitation Assessment
Program. Washington, pp. 27-177 to 27-195. 1991.
5. Cicchetti, C. and A. Freeman. "Option Demand and Consumer Surplus: Further
Comment," Quarterly Journal of Economics. Vol. 85, pp. 528-539. 1971.
6. Cummings, R., D. Brookshire, and W. Schultze. Valuing Environmental Goods: An
Assessment of the Contingent Valuation Method. Rowman and Allanheld. Totowa, New
Jersey. 1986.
7. Desvousges, W., V.K. Smith, and A. Fisher. "Option Price Estimates for Water Quality
Improvements: A Contingent Valuation Study for the Monongahela River," Journal of
Environmental Economics and Management. Vol. 14, pp. 248-267. 1987.
8. Doyle, J., S. Elliot, G. McClelland, and W. Schultze. 1991. Valuing Benefits of Ground
Water Cleanup: Interim Report. Center for Economic Analysis, Dept. of Economics, Univ.
of Colorado. Boulder, CO. January 1991.
9. Edwards, S. F. "Option Prices for Ground Water Protection," Journal of Environmental
Economics and Management. Vol. 15, pp. 465-487. 1988.
10. Fisher, A. and R. Raucher. "Intrinsic Benefits of Improved Water Quality: Conceptual and
Empirical Perspectives," Advances in Applied Econometrics. JAI Press, Greenwich. Vol. 4.
1984.
11. Freeman, A. "The Quasi-Option Value of Irreversible Development," Journal of
Environmental Economics and Management. Vol. 11, pp. 292-295. 1984.
12. Freeman, A. "The Sign and Size of Option Value," Land Economics. Vol. 60, pp. 1-13.
1984.
13. Freeman, A. "Non-Use Values in Natural Resource Damage Assessment," Paper prepared
for the Conference on Assessing Natural Resource Damages, June 16-17, 1988. Resources
for the Future, Washington DC. 1988.
GW4-23
-------
Session GW-4
Crutchfield
14. Krutilla, J. "Conservation Reconsidered," American Economic Review. Vol. 57, pp. 777-
786. 1967.
15. Krutilla, J., C. Cicchetti, A. Freeman, and C. Russell. "Observations on the Economics of
Irreplaceable Assets," Environmental Quality Analysis: Theory and Method in the Social
Sciences. The Johns Hopkins University Press, Baltimore, MD. 1972.
16. Madariaga, B. and K. McConnell. "Exploring Existence Value," Water Resources Research.
Vol. 23, No. 5, pp. 936-942. May 1987.
17. McConnell, K. "Existence and Bequest Values," Managing Ah" Quality and Scenic Resources
in National Parks and Wilderness Areas. Westview Press, Boulder, CO. pp. 254-264.
18. Nielsen, E.G. and L.K. Lee. The Magnitude and Costs of Ground Water Contamination
from Agricultural Chemicals. UDSA/ERS Agricultural Economics Report No. 576.
October 1987.
19. Shultz, S. and B. Lindsay. "The Willingness to Pay for Ground Water Protection," Water
Resources Research. Vol. 26, pp. 1869-1875. 1990.
20. Smith, V.K. "Nonuse Values in Benefit Cost Analysis," Southern Economic Journal.
Vol. 54, No. 1. July 1987.
21. Smith, V.K. "Uncertainty, Benefit-Cost Analysis, and the Treatment of Option Value,"
Journal of Environmental Economics and Management. Vol. 14, pp. 283-292. 1987.
22. Sun, J. "Economic Analysis of Ground Water Pollution by Agricultural Chemicals,"
Unpublished Master's Thesis, Dept. of Agricultural and Applied Economics, The University
of Georgia. 1990.
23. U.S. Environmental Protection Agency. National Pesticide Survey Phase 1 Report.
Washington,.DC, EPA 570/9-90-015. 1990.
24. Weisbrod, B. "Collective Consumption Services of Individual Consumption Goods,"
Quarterly Journal of Economics, pp. 471-77. August 1964,
25. Williams, W.M. et. al. Pesticides in Ground Water Database - 1988 Interim Report.
December 1988.
GW4-24
-------
Session GW-4
Crutchfield
APPENDIX I
DRAFT HEALTH AND WATER QUALITY CV QUESTIONS
The following questions will be randomized in some fashion. Specifically, the values of XXX, Y, and Z will vary
across individuals in a random fashion.
NOW, FOR A CHANGE OF PACE, I WOULD LIKE TO ASK YOU SEVERAL QUESTIONS
REGARDING YOUR HOME DRINKING WATER SUPPLY.
(read this always)
Water from your tap, whatever its source, can sometimes contain small amounts of impurities. Some
impurities in water give it an unpleasant taste (and you might have installed a filter to make your water taste
better). Other chemicals or impurities, that you cannot taste, may be a risk to your familys health. For
example, nitrates or pesticides, when they are present in drinking water at high enough levels, increase the
risk of some forms of cancer, infant mortality, or other health problems. The EPA and other health officials
regularly publish standards of exposure to chemicals in drinking water which are thought to be reasonably
safe. Now for the question...
220A. Suppose your drinking water supply was found to have chemicals in it at some level above
what is considered "safe", but which placed your family at no immediate health risk. That is,
drinking your water would not make you ill right away, but over several years might raise
your risk of becoming ill. Also, suppose that these chemicals could be reduced to levels
health experts consider safe, but not eliminated, by installing a filter system in your home.
If this filter system cost $XXX per month, would you purchase it?
1. YES -- skip to 220B A. Refused -- skip to 221
2. NO -- skip to 220C B. Don't know -- 221
C. Not Ascertained -- skip to 221
220B If the filter cost was $XXX + Z per month, would you (still) buy it?
1. YES - skip to 220E A. Refused -- skip to 221
2. NO - skip to 220E B. Don't know --221
C. Not Ascertained -- skip to 221
220C If the filter cost was lowered to $XXX-Z per month, would you buy it?
1. YES -- skip to 220E A. Refused -- skip to 221
2. NO -- skip to 220D B. Don't know -- 221
C. Not Ascertained -- skip to 221
220D Why did you answer NO to this question? I will read a list of 5 reasons, please choose the
most appropriate (if none match, write down OTHER).
1) I do NOT expect to receive ENOUGH BENEFITS from the proposed filter system.
- skip to 220E
or
2) I CANNOT AFFORD HIGHER water treatment COSTS at this time.
- skip to 220e
or
3) Clean water IS MY RIGHT and it is UNFAIR to ask me TO PAY.
- skip to 221
4) The government should pay for this, even if it means higher taxes.
- skip to 221
GW4-25
-------
Session GW-4
Crutchfield
5) I would rather buy bottled water than install this filter.
~ skip to 221
A) Refused -- skip to 221
B) Don't know -- skip to 221
C) Not Ascertained - skip to 221
D) OTHER (specify for pre-test)
220E Suppose instead that a more advanced filter system could eliminate these chemicals entirely,
but would cost $XXX + YYY per month. Would you pay this amount to eliminate these
potentially harmful chemicals entirely?
1. YES -- skip to 220F A. Refused -- skip to 221
2. NO -- skip to 220G B. Don't know - 221
C. Not Ascertained - skip to 221
220F If this advanced filter cost $XXX +YYY+Z per month, would you (still) buy it?
1. YES -- skip to 221 A. Refused -- skip to 221
2. NO -- skip to 221 B. Don't know - 221
C. Not Ascertained - skip to 221
220G If this advanced filter cost was lowered to $XXX+YYY-Z per month, would you buy it?
1. YES -- skip to 221 A. Refused -- skip to 221
2. NO -- skip to 220H B. Don't know -- 221
C. Not Ascertained ~ skip to 221
220H Why did you answer NO to this question? I will read a list of 5 reasons, please choose the
most appropriate (if none match, write down OTHER).
1) I do NOT expect to receive ENOUGH BENEFITS from the proposed filter system.
or
2) I CANNOT AFFORD this HIGHER water treatment COST at this time.
or
3) Clean water IS MY RIGHT and it is UNFAIR to ask me TO PAY.
4) The government should pay for this, even if it means higher taxes.
5) I would rather buy bottled water than install this filter.
A) Refused
B) Don't know
C) Not Ascertained
D) OTHER
221 Is your home connected to a municipal water supply?
1. YES A. Refused -- skip to 230
2. NO - skip to 221B B. Don't know --
C. Not Ascertained ~ skip to 230
221A Is this supply from ground water?
1. YES A. Refused -- skip to 230
2. NO B. Don't know --
C. Not Ascertained -- skip to 230
GW4-26
-------
Session GW-4
Crutchfield
221B Do you get your drinking water from a private well?
1. YES A. Refused -- skip to 230
2. NO B. Don't know --
C. Not Ascertained - skip to 230
221C Do you currently use a private water treatment system?
1. YES A. Refused - skip to 230
2. NO - skip to 230 B. Don't know -
C. Not Ascertained - skip to 230
221D What type of system are you using (interviewer match answer to list below)!
1) activated carbon filter
2) reverse osmosis
3) other filter
4) systems which vent water to steam
5) aerator faucets
6) purchase bottled water -
7) add bleach to water
8) boil water
9) Other.
A. Refused-skip to 230
B. Don't know
C. Not Ascertained
221E Is the primary reason you use a private water treatment system:
1) To improve taste or odor of the water
or
2) For safety/health reasons
or
3) Other reasons.
A. Refused
B. Don't know
C. Not Ascertained
GW4-27
-------
Session GW-4
Cropper
Research Issues in Valuing Ground Water
Maureen L. Cropper
Professor, University of Maryland and
Senior Fellow, Resources for the Future
ABSTRACT
This paper focuses on the relationship between benefit estimates that would be useful to EPA in issuing
ground water regulations and current research on the value of unpolluted ground water. The main issue
concerns what is to be valued. In the area of current use values, one approach would be to translate ground
water pollution into health risks (e.g., number of cancer cases) and have consumers value these health
endpoints. This approach could be implemented using contingent valuation or using averting expenditures.
In the latter case it is important to know consumers' perceptions of the risk reduction they are achieving
through averting behavior.
1.0 EPA DECISIONS AFFECTING GROUND WATER
In valuing ground water, how should a research effort be directed if it is to be useful to the
Environmental Protection Agency (EPA)? In answering this question I will take the perspective of an
economist who believes that benefit-cost analyses can be a useful aid to regulation. This implies that effort
should be put into having people assign dollar values to various aspects of ground water regulations. The
regulations on which I will focus are those issued under the Resource Conservation and Recovery Act
(RCRA) and the Comprehensive Environmental Response, Compensation and Liability Act (CERCLA).
Regulations issued under RCRA and CERCLA affect ground water in two ways. Corrective
action regulations under RCRA and the choice of remedial actions at Superfund sites under CERCLA
determine how much existing ground water contamination to remediate. RCRA design standards for
landfills and the choice of cleanup strategy for contaminated soils at Superfund sites reduce the probability of
future ground water contamination.
As far as cleanup of existing contamination is concerned, there are basically two decisions
that must be made:
1. How extensively should one clean up contamination? This is equivalent to asking
what is an acceptable maximum contaminant level (MCL) for various pollutants in
ground water.
2. How permanently should ground water be cleaned up? This is equivalent to asking
whether (in addition to point-of-use treatment) there should be some form of in
situ treatment, or, whether ground water should be pumped and treated hi order to
reduce the size of the plume of contamination.
GW4-28
-------
Session GW-4
Cropper
2.0 WHAT IS TO BE VALUED?
What decisionmakers need to know to answer the first question is the value associated with
lower MCLs. Part of the value of lower MCLs can be expressed in terms of reductions in risk of death. It
is these reductions in mortality risk, rather than reductions in MCLs that consumers should be asked to
value. Although reductions in risk of death are unfamiliar to most people, they are probably easier to
comprehend than reductions hi pollutant concentrations. Secondly, the value of risk reductions are
transferr able-they can be used in a variety of circumstances, assuming that scientists can supply the link
between MCLs and risk levels.
The value of increased permanence of treatment reflects the value of reducing MCLs for
future generations, as well as the value of cleaning up ground water for non-consumptive uses (the so-called
"non use" value of ground water). The value of the latter is essentially what people would be willing to pay
for a pump and treat strategy at a site versus what they would be willing to pay for treatment of water
withdrawn from the aquifer.
3.0 ALTERNATIVE VALUATION TECHNIQUES
3.1 Techniques Used to Value Cleanup of Existing Contamination
How successful are the various valuation techniques used by economists likely to be hi
measuring these two components of value? There is a large economics literature on the value of reductions
hi risk of death;1 however, most of this is based on voluntarily assumed risks. To value reductions in
involuntarily assumed risk of death one could use either an averting behavior approach or a contingent
valuation approach.3 The averting behavior approach uses expenditures (e.g., on bottled water) to avoid
exposure to contamination to value clean ground water. The contingent valuation approach asks people
whether they would pay a stated amount for a given risk reduction.
The value of the averting behavior approach is that it is based on actual, rather than
hypothetical, behavior; however, for an averting behavior study to be useful in valuing risk reductions, one
must know how much increased safety people thought they were buying by purchasing bottled water. This
entails the difficult task of communicating with people about the size of small risk reductions. This task,
however, is also necessary, if one is to use the contingent valuation approach.
Another possible approach to valuing reductions in health risks is to look at the premium
people are willing to pay to move away from hazardous waste sites, or, equivalently, the loss hi value of
houses located near such sites.4 As with the averting behavior approach, the problem here is hi knowing
what the decrease hi property values reflects. What gain in safety do people think they are purchasing by
moving away from a site?
The task of valuing increased permanence of ground water cleanup has been attempted by
Schultze5 who asked different groups of individuals what they would be willing to pay for (a) point-of-use
treatment; (b) point-of-use treatment plus pump-and-treat to contain the plume; (c) pump-and-treat to
remove all contamination. The difference in average willingness to pay between (a) and (b) and between (b)
and (c) indicates the value of more permanent treatment.
Studies of this type are essential if EPA is to determine whether the extremely high cost of
the protracted (e.g., 30-year) pump-and-treat remedies chosen at some Superfund sites is worthwhile. For
these studies to be useful, however, EPA must know how willingness to pay varies with (1) baseline risk
(contaminant level) at the site; (2) the size (in acres) of the aquifer contaminated; (3) the likelihood of
surface water contamination or other adverse ecological consequences.
GW4-29
-------
Session GW-4
Cropper
Techniques Used to Value Risk of Future Contamination
The problem of valuing reductions in the risk of future ground water contamination, like the
problem of valuing increased permanence of cleanup, is probably best approached through the use of
contingent valuation. Here the researcher must specify both parts of the commodity being valued: the
change in the risk of adverse consequences and the nature of the adverse consequences themselves. To date,
two approaches have been taken. Smith and Desvousges6 value the probability of exposure to hazardous
waste as well as the probability of death, assuming that exposure occurs. Bergstrom values a reduction the
probability of "unsafe" ground water.
Although individuals clearly had difficulties dealing with compound probabilities in,6 the
notion that ground water contamination can be classified as "safe" versus "unsafe" seems of little use in
issuing regulations unless one (artificially) separates the problem of the optimal MCL from the issue of
valuing the risk of contamination. This implies that success in valuing ground water hinges on the ability of
researchers to communicate risk changes to consumers and to have consumers value them.
To some people, this may appear a hopeless task. Because risks are difficult to
communicate, it may seem easier simply to have people vote on whether to undertake ground water
cleanups, based solely on a physical description of the contamination. This, however, has one pitfall. If
people are allowed to imagine the risks associated with contaminated ground water, then policy may be
based on erroneous risk perceptions. Even if this is not a problem, the contingent valuation study has an
advantage. If people correctly perceive the risks associated with ground water contamination, a contingent
valuation study should, in fact, come to the same conclusion as a referendum. The advantage of the former,
however, is that, by explicitly valuing risk reductions, the results can be used in a variety of situations.
4.0 REFERENCES
1. Cropper, M. and A.M. Freeman III. "Environmental Health Effects," Measuring the
Demand for Environmental Quality. Amsterdam, North-Holland. 1991.
2. Abdalla, C. "Avoidance Costs and Ground Water Values: Results of Two Empirical
Applications," Paper presented at the conference on Clean Water and the American
Economy, Arlington, VA, Washington, D.C. October 20, 1992.
3. Mitchell, R.C. and R.T. Carson. "Valuing Drinking Water Risk Reductions Using the
Contingent Valuation Method: A Methodological Study of Risks from THM and Giardia,"
Draft Report to the U.S. Environmental Protection Agency, Washington, D.C. 1986.
4. Michaels, G. "When the Home is No Longer A Castle: Ignoring the Economic Value of
Ground Water Contamination from Residential Property Values," Paper presented at the
conference on Clean Water and the American Economy, Arlington, VA. Washington, D.C.
October 20, 1992.
5. Schultze, W. "Using Contingent Valuation to Measure Bequest and Existence Values for
Ground Water Cleanup," Paper presented at the conference on Clean Water and the
American Economy, Arlington, VA. Washington, D.C. October 20, 1992.
6. Smith, V.K. and W.H. Desvousges. "An empirical analysis of the Economic Value of Risk
Changes," Journal of Political Economy. No. 95. pp. 89-114. 1987.
7. Bergstrom, J. "Benefits of Protecting Ground Water from Agricultural Chemical
Contamination," Paper presented at the conference on Clean Water and the American
Economy, Arlington, VA. Washington, D.C. October 20, 1992.
GW4-30
-------
Session GW-4
Howarth
Environmental Risks and Future Generations;
Criteria for Public Policy*
Richard B. Howarth
Energy and Environment Division
Lawrence Berkeley Laboratory
ABSTRACT
This paper examines alternative normative approaches to the policy challenges posed by
long-term environmental problems such as toxic and radioactive waste disposal, stratospheric ozone
depletion, and climate change. The paper argues that cost-benefit analysis is limited in its ability to handle
the issues of intergenerational equity and uncertainty that are intrinsic to such problems. Also considered is
the precautionary principle, which holds that policies should seek to reduce threats to the welfare of future
generations if the costs of doing so would not significantly reduce the subjective well-being of existing
persons. Although the precautionary principle depends on an explicit value judgement, it yields a policy
criterion that is operationally decisive under a wide array of circumstances.
1.0 INTRODUCTION
A broad class of environmental problems is characterized by the asymmetric distribution of
benefits and impacts over time. Activities that generate toxic and radioactive substances, greenhouse gases,
and other environmental insults yield perceived benefits to todays economy. But while wastes may be
contained in steel drums, storage tanks, or underground depositories for some period of time, the possibility
of their eventual release to the soil or water imperils persons who will live decades, centuries, or even
millennia in the future. Similarly, greenhouse gas emissions have few immediate effects, but cumulative
emissions threaten the long-term stability of the global climate system with potentially far-reaching
implications for human and ecological systems.
The impacts of such long-term environmental threats are highly uncertain. Many products
of modern technology do not exist apart from their manufacture by humans. Others, while naturally
occurring, are today released by human activities at rates rivalling or exceeding the assimilation capacity of
nature. As environmental scientist Wallace Broecker1 pointed out, "[t]he inhabitants of planet Earth are
quietly conducting a gigantic environmental experiment." Because we are perturbing the environment in a
manner lying outside the range of historical experience or the capabilities of laboratory simulation, there can
be no certainty regarding the impacts of climate change, the safety of hazardous waste storage, and so forth.
The example of stratospheric ozone depletion is instructive on this point. In the 1970s,
simple calculations based on laboratory measurements showed that chlorofluorocarbons had the potential to
deplete the ozone layer, but the magnitude of the effect was generally taken to be small. The discovery of
the Antarctic ozone hole in the early 1980s at first baffled scientists. Subsequent research established that
ice crystals embodied in certain Antarctic clouds greatly accelerate the rate of ozone depletion. The ozone
layer thus turned out to be much less stable than the early calculations led us to believe.
Approaching the policy challenges posed by long-term environmental problems is
controversial both in theory and in practice. Some argue that policies should be designed to equate the
marginal costs and benefits of pollution abatement, measured in monetary terms. Others argue that the
imposition of long-term environmental risks is morally unacceptable because it threatens the welfare of
aThis work was sponsored by the Stockholm Environment Institute through the U.S. Dept. of Energy under
Contract No. DE-AC03-76SF00098.
GW4-31
-------
Session GW-4
Howarth
future generations. This tension is clearly evident in U.S. law. The National Environmental Policy Act, for
example, recognizes "the responsibilities of each generation as trustee of the environment for succeeding
generations," while the Clean Air Act explicitly forbids the consideration of cost-benefit criteria in
promulgating air quality regulations. Executive Order 12291, on the other hand, requires that all federal
regulations be evaluated using cost-benefit techniques unless countermanded by statute.2
Untangling the issues behind this controversy is the focus of this paper. The paper begins
with a discussion of cost-benefit analysis, outlining two characteristics that constrain its usefulness in the
analysis of long-term environmental problems. First, the approach is blind to the distribution of impacts
between social groups and between present and future generations. To the extent that issues of
distributional equity are important to decision makers, this necessitates the use of explicit ethical criteria in
policy formulation and evaluation. Second, cost-benefit techniques are generally ill-suited to the analysis of
problems characterized by substantial uncertainty. In theory, cost-benefit analysis under uncertainty is a
simple extension of well-established methods. In practice, the information requirements are often beyond the
means of practical implementation.
As an alternative to cost-benefit analysis, the paper explores the implications of the
precautionary principle as a normative approach to long-term environmental management under uncertainty.
The precautionary principle is derived from the concept of sustainable development, and holds that policies
should seek to reduce threats to future welfare if the costs of doing so would not significantly reduce the
subjective well-being of present or future persons. This principle, like other normative criteria for use in
policy analysis, rests on a particular value judgement. If one accepts this value judgement as reasonable,
then one is left with a policy criterion that is operationally decisive under a wide array of circumstances.
2.0 THE THEORY OF COST-BENEFIT ANALYSIS
Cost-benefit analysis is rooted in a simple but compelling ethical proposition. By the
doctrine of Pareto efficiency, a proposed policy change will lead to an improvement in social conditions if it
benefits at least some members of society while leaving none worse off. Actual policy changes generally
benefit some individuals but harm others, so this maxim would appear on the surface to have limited
relevance to the real world. Suppose we define the net monetary benefit accruing to each individual as his or
her net willingness to pay for a proposed policy change. If we assume that people are the best judges of
their own well-being and that they are economically rational, a policy change will improve their welfare if
they would be willing to pay a positive sum of money to put it into effect. Conversely, they would be injured
if they would be willing to pay to prevent implementation of the policy. If the summed positive benefits
accruing to the winners are greater than the summed "costs" or negative benefits incurred by the losers, then
in principle the winners could compensate the losers so that the welfare of all individuals could be improved.
Policy proposals that satisfy this standard, sometimes termed the Kaldor-Hicks criterion, are termed potential
Pareto improvements and may in principle be identified using cost-benefit analysis.
A broad range of techniques have been devised to measure the net willingness to pay for
proposed policy changes. ' In the simplest case, net benefits are measured by multiplying the change in the
availability of each affected good by its price, assuming that no price changes are induced by the policy.
Where the change is nonmarginal so that not only quantities but also prices are affected, the appropriate
indicator is the change in "social surplus," approximated in competitive markets by the area bounded by the
market supply and demand functions between the initial and final quantities of the good.5
A pervasive problem in cost-benefit analysis is the aggregation of costs and benefits that
accrue at different points in time. Generally speaking, future benefits are worth less than those of the
present since a dollar today may be invested to yield 1.03 dollars next year given a 3% interest rate. In
neoclassical models of intertemporal equilibrium under perfect foresight, the interest rate constitutes a
measure of an individual's marginal preference for consumption in sequential periods. This fact does not
imply that people prefer the present to the future in any abstract sense-only that they optimize their
consumption streams in a world of investment opportunities. In this sense, the discount rate is simply the
price of future consumption relative to present consumption.
GW4-32
-------
Session GW-4
Howarth
To express present and future benefits hi comparable present-value units, net benefits that
are realized t periods from the present are discounted by the factor
*t = TT-J- W
ri
where rt is the interest rate at date t and S 0 = !• 1° tne special case where the interest rate is constant over
time so that rt = r, this formula reduces to the more familiar 1/(1 + r)'. Suppose that Ct and Bt are the
flows of monetary costs and benefits realized at time t as a result of the proposed policy change. Then the
net present value (NPV) of the net benefits yielded by the proposed policy change is given by
NPV=£5t(H-q) (2)
t-0
where the current date is normalized to t = 0 and T is the final date at which the policy has economic
impacts. If this quantity is positive, then the policy change constitutes a potential Pareto improvement and is
said to yield net positive benefits in the sense that the policy could in principle be implemented along with
appropriate income transfers so that all members of society would be rendered better off. It is
well-recognized, however, that a potential Pareto improvement need not constitute an actual Pareto
improvement. If policy implementation benefits some individuals at the expense of others and no
compensation follows, the logic supporting cost-benefit analysis breaks down. Potential Pareto improvements
constitute unambiguous opportunities for improved social welfare only if the "losers" are duly compensated.
2.1 Cost-Benefit Analysis and Distributional Equity
A distinguishing characteristic of cost-benefit analysis is its marriage to the baseline. All of
the variables that go into a cost-benefit calculation-the cost of pollution abatement, the associated
environmental benefits, and the discount rate—are reflections of anticipated economic conditions. The future
path of the economy is not, however, fixed in stone but is instead of matter of collective choice. Should we
as a society use the resources at our disposal to maximize our own selfish gratification without regard to the
welfare of future generations? Should we act so as to ensure that the life opportunities of our children and
grandchildren are equivalent to or better than our own? Either choice is possible and either may be pursued
with consummate economic efficiency. Yet the efficient balance between the costs and benefits of pollution
abatement might vary sharply under the two scenarios.
Consider, for example, the case of a long-lived pollutant generated by current human
activities. Suppose for simplicity that the future costs imposed by the pollutant vary in linear proportion with
economic activity. Then strong economic growth would raise pollution damages relative to a low-growth
scenario at each point in time. As we have seen, the discount rate appropriate for use in cost-benefit
analysis is equal to the marginal return on capital investment in the absence of market distortions. Economic
growth is fueled by capital investment, with the rate of capital accumulation involving an equity decision
concerning the level of wealth we wish to transfer to future generations. Increased accumulation implies a
decrease in the marginal return on investment and hence a reduction in the social discount rate. Together,
higher impacts and lower discount rates imply that it would be efficient to abate pollution more aggressively
in a high-growth world than in a low-growth alternative.
This argument rests on particular theoretical and empirical assumptions and is rather
informal in character. It is possible, however, to illustrate similar results using formal models rooted in the
theory of intertemporal general equilibrium. Focusing on the issue of climate change, for example, Howarth
and Norgaard showed that cost-benefit techniques may be used to identify efficient greenhouse gas
emissions profiles in a hypothetical overlapping generations economy. The efficient outcome, however,
depends strongly on the degree of caring for the future, with an efficient world of deplorable living standards
and high pollutant levels for future generations ours for the choosing should we so desire.
GW4-33
-------
Session GW-4
Howarth
While the details of the Howarth-Norgaard model need not concern us here, a review of its
results provides some insight into the subject under discussion. Figure 1 shows the levels of key economic
variables-per capita consumption, the capital stock, greenhouse gas concentrations, and the social discount
rate-for two model runs. The "impoverished future" case assumes an ethical framework in which present
society cares little for posterity and thus depletes capital assets and adds substantially to the stock of
greenhouse gases. The "sustainable future," hi contrast, assumes that the present generation preserves capital
goods and environmental quality for the sake of future generations. In each case, cost-benefit criteria are
applied to identify an efficient greenhouse gas emissions profile. The two differ hi the transfers of assets that
are effected from one generation to the next, equivalent to transfers of wealth from the rich to the poor
motivated by concerns about social justice.
The use of cost-benefit procedures will lead to an efficient response to long-term
environmental threats only if the analyst correctly anticipates the future course of the economy. This
represents a logical paradox since the "all else equal" assumptions of partial analysis are of limited relevance
in an environment where all policy variables are subject to simultaneous choice. As Dasgupta and Heal
pointed out, an economic future "can be intertemporally efficient and yet be perfectly ghastly" if it denies
future generations the physical and cultural conditions required to sustain a satisfactory way of life. These
facts should give us pause for thought about the use of cost-benefit analysis to identify an "optimal" response
to problems such as waste storage, ozone depletion, and climate change in a world where issues of
intergenerational equity are perceived to be at stake.
It is sometimes argued that scientific and technical progress are paving the way to a world
of future abundance, obviating the need to consider questions of intergenerational equity in the analysis and
promulgation of public policy. Indeed, the centuries since the Industrial Revolution have^been marked by
profound improvements in living standards driven by fundamental transformations in the interrelationships
between technology, social institutions, and the natural environment. But trends are not destiny, and one
cannot safely assume that conditions will improve in the future simply because they have unproved hi the
past. Some now argue that the trend towards economic progress has already reversed and that today's young
people will be unable to match the standard of living achieved by their parents in the absence of policy
intervention.8 The question to ask is as follows: Is the present generation contributing to the technological
base and preserving the capital and natural assets required to sustain the future welfare hi light of anticipated
technological progress and emerging environmental constraints?
Ours is the power to confer a world of poverty or abundance to the members of future
generations, and there is no guarantee that events will turn out favorably in the absence of careful planning
regulated by the adoption of suitable planning criteria.
22 Cost-Benefit Analysis and Equity Between Contemporaries
The focus of this paper is on the issue of intergenerational equity as it relates to
environmental policy analysis. This focus, while helpful for purposes of exposition, is in truth a bit artificial
in character. Environmental impacts, after all, are not distributed uniformly amongst contemporaries.
Typically, it is the weak and the vulnerable who bear the largest burden. For example, exposure to
hazardous waste will fall disproportionately on the poor who must live next to waste storage facilities for
want of the resources necessary to rent a home in a more desirable neighborhood. Similarly, the individuals
most adversely affected by climate change are likely to be residents of low-income nations lacking the means
to adapt favorably to changing climatic conditions.
Conventional cost-benefit techniques place equal weight on net monetary benefits that
accrue to contemporaries regardless of their relative welfare. This runs against our moral intuition, for many
would argue that a dollar spent on the poor yields benefits of greater moral worth than a dollar spent on the
rich. In theory, issues of equity could be redressed through the transfer of wealth from rich to poor. In
practice, however, disparities of wealth are likely to persist; is it then acceptable to impose environmental
burdens on the weak so that the affluent may enjoy marginal benefits? Although such transfers of wealth
from poor to rich might pass the test of Pareto efficiency, they are difficult to defend on moral grounds.
GW4-34
-------
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
Session GW-4
Howarth
Figure 1: Alternative Future Worlds
(a) Impoverished Future
aOMOKiiaCltXliqOICK^^
4-
Consumption
Capital Stock
Greenhouse Gases
Discount Rate
5 10 15 20 25 30 35 40
Generations from Present
(b) Sustainable Future
H-H-
-I I It I I I I I I If' I I I I I I I I I I I I I I I I I I I I I
Consumption
Capital Stock
Greenhouse Gases
Discount Rate
0 5 10 15 20 25 30 35 40
Generations from Present
GW4-35
-------
Session GW-4
Howarth
In principle, we need draw no distinction between fairness amongst contemporaries and
fairness between present and future generations. Principles of justice between contemporaries logically
define obligations to even the distant future.9 In practice, this means that we need to focus not only on the
timing of impacts but also on their distribution between social groups.
23 Cost-Benefit Analysis and Uncertainty
Analyses of the potential costs and benefits of proposed environmental policies often focus
on expected outcomes, averaging across low- and high-impact scenarios to obtain an estimate of the most
likely sequence of events. In the face of substantial uncertainties, such a focus is not entirely appropriate.
Our intuition informs us that fire insurance is a good investment even though we hope and expect that our
homes will never burn down. Put another way, individuals will often give up expected benefits to protect
themselves against the possibility of entailing large losses.
One approach to cost-benefit analysis under uncertainty is to use ad hoc procedures to
adjust expected outcomes to account for risk. A standard argument is that individuals demand higher
expected rates of return on investments yielding risky benefit streams in comparison with secure investments
such as long-term government bonds. Thus cost-benefit analysts sometimes apply high discount rates in
evaluating uncertain projects. While such an approach is simple to apply in practice, in theory it is rather
objectionable.10 Theory informs us that a rational investor will demand a high expected rate of return on an
uncertain investment if its returns are positively correlated with the return on her/his overall investment
portfolio. Conversely, she/he will accept comparatively low (or even negative) expected returns on assets
that provide insurance by yielding high returns when the market as a whole turns sour.
To illustrate the difficulties inherent in the problem, it is helpful to outline the formal
criterion used to determine whether a policy offers a potential Pareto improvement when its outcome is
uncertain. Suppose that there are n(t) possible outcomes or "states of nature" at date t denoted sti for
i=l,...,n(t). The probability of each state is Pr(sti). If the policy yields the net benefit Bti - Cti under state sti
at date t, the policy yields a potential Pareto improvement if the present-value expression
T n(t)
NPV = £E *(%)**<%-qi) (3)
t=0 M
is greater than zero.11 The discount factor 6 ti deserves special comment. In general, this factor varies
across time and states of nature, accounting simultaneously for individual preferences concerning both time
and risk. Each contingent future is linked to the present by its own state-contingent discount factor. The
discount factor depends on individuals' risk aversion and on their relative well-being at sequential dates and
under alternative states of nature.
It is clear that enormous quantities of information would be required to rigorously evaluate
the costs and benefits of proposed policy interventions given substantial uncertainty. Consider the case of a
long-lived pollutant with the potential to cause catastrophic harm to members of future generations. To
decide on an efficient level of pollution abatement, we would need to know the complete range of possible
future states, including their statistical probability, environmental impacts, and implications for human
welfare. We would need to gauge social preferences regarding time and risk, even for low-probability,
extreme outcomes for which we have little hard information to fall back on. It is not difficult to see that this
approach is generally inoperational-we cannot with confidence identify efficient policy responses to long-term
environmental problems where uncertainties loom large.
Where does that leave cost-benefit analysis as an approach to environmental policy? We
can use crude information to get some feeling for the expected impacts of environmental insults as well as
the probability of extreme change. We can reasonably speculate that society would be willing to spend extra
resources to mitigate the threat of potentially catastrophic risks. But the appropriate sum to pay is beyond
GW4-36
-------
Session GW-4
Howarth
the reach of economic analysis and thus depends on the exercise of raw value judgements regarding what is
acceptable and what is not.
3.0 SUSTAINABILITY AND THE PRECAUTIONARY PRINCIPLE
The class of problems under discussion has potentially far-reaching but uncertain
consequences for the distribution of welfare between present and future generations, yet cost-benefit
techniques are inherently ill-equipped to address issues of equity and uncertainty. How then should we
proceed in the formulation and evaluation of policy? One possibility is to posit the existence of a social
welfare function as a means of comparing and evaluating alternative strategies based on the comparative
welfare of present and future generations across the complete range of possible outcomes. In principle, a
social welfare function would simultaneously cope with questions of equity and uncertainty by reducing social
values to a single, well-defined criterion. In actuality, efforts to define an appropriate welfare function have
proved inoperational for well-known theoretical and practical reasons.12 Even if the presumed welfare
function were within our grasp, its application to compare alternative policies would run up against the same
information requirements that confound cost-benefit analysis under uncertainty.
We are left then to identify alternative criteria that capture prevailing notions of
intergenerational justice under uncertainty. To gain some insight into this problem, it is important to note
that the notion of intergenerational equity as it is usually put forth in public debates over environmental
policy takes the form of a constraint on the range of outcomes that are considered ethically permissible
rather than a utilitarian definition of "optimal" distribution. As the criterion is usually stated, economic
development should be sustainable in the sense that the utilization of natural resources and the environment
by the present generation does not jeopardize the ability of future generations to enjoy a favorable standard
of living.
A number of definitions of sustainability and sustainable development have appeared in the
literature. Consider, for example, the following selections:
"The sustainability criterion suggests that, at a minimum, future generations should be left
no worse off than current generations."13
"Sustainable development is development that meets the needs of the present without
compromising the ability of future generations to meet their own needs."14
"A sustainable society is one that satisfies its needs without jeopardizing the prospects of
future generations. Inherent in this definition is the responsibility of each generation to
ensure that the next one inherits an undiminished natural and economic endowment."15
These definitions are rooted in the common principle that present and future generations are ethically
equivalent although they are not contiguous in time. Hence, morality requires that members of future
generations have equal or better opportunities than the present generation to live the good life in the same
sense that it mandates an equitable distribution amongst the current generation. More so, in fact, since one
might argue that while some degree of distributional inequality within a generation might be justified by the
relative merits of individuals—the rich may have earned their wealth while the poor may have brought poverty
upon themselves-it is difficult to argue that future generations are as a group less deserving than the present.
To argue otherwise would be to discriminate against future generations based on the arbitrary happenstance
of their birth dates.
Some philosophers, on the other hand, maintain that the present is in general under no
obligation to provide a resource-rich world to future generations, or at least that such obligations are very
weak, Schwartz16'17 for example, has argued that even minor policy changes intended to improve the lot of
future generations would change not only the welfare but also the composition of future generations. Hence
we are unable to affect the living standards of a well-defined set of future individuals; instead, we are
choosing whether to bring relatively rich or relatively poor individuals into existence. If we take as our
GW4-37
-------
Session GW-4
Howarth
assumption that an action is morally mandated only if it benefits some individual who will actually exist, then
this argument seems to force the conclusion that beneficence to future generations is not morally required
unless the future world is so poor that the lives of future generations are not worth living.
Does this argument undermine the ethical basis of the sustainability criterion? Suppose that
we define distributional equity as follows: All individuals, both present and future, should have an equal
opportunity to pursue their own welfare. According to this criterion, a nonsustainable development program
may harm no particular future individual but nonetheless be morally wrong on the basis that it gives rise to
an unjust welfare distribution.18'19'20'21'22
Schwartz's line of reasoning is open to another powerful critique. Children are born into
this world helpless but for the benevolence of their parents and society generally. Each generation and the
next overlap in time, and from a parent's perspective, children are not future contingencies but rather facts of
day-to-day existence. Most would agree that parents are under a strong obligation to provide their children
with life opportunities at least equivalent to their own. For parents and their living offspring are morally
distinct only in the happenstance of their birthdates, and it would be unjust for parents to pursue their own
selfish interests at the expense of their children simply because their age and familial authority empowered
them to do so.
Although the identities of unborn persons remain undetermined, our children will be
obligated to their children once they are born and become flesh and blood. Thus our actions must ensure
our children a favorable existence while permitting them to honor their obligation to their offspring. By
logical extension, this argument defines a chain of obligation between the present and the indefinite future to
ensure that living standards are nondeclining from generation to generation. We owe it to our children, who
will owe it to their children, and so on as far as the mind can see.9
But even if sustainability is not deducible from prior ethical principles, it is nonetheless of
direct policy relevance to the extent that it reflects the distributional values of the current generation.
Indeed, the available evidence as reflected by the proclamations of politicians and related indicators of public
opinion points to a high degree of concern in the body politic for the welfare of future generations.
The success of the sustainability criterion as a guide to policy analysis depends critically on
the translation of these general precepts into operational planning criteria. But while there may be
agreement on underlying values, there is considerably less on the implications of these values for
intertemporal planning. As Lele23 pointed out, the term "sustainable development" will devolve into a
meaningless catch-phrase unless it is carefully and operationally defined.
Neoclassical economists have interpreted sustainability as a technical requirement that the
utility or welfare of successive generations should be no lower than that of their predecessors. Pezzey,24 for
example, explored the implications of the sustainability criterion for simple models of intertemporal
development, reaching the conclusion that sustainability is a constraint that allows some degree of flexibility
in intertemporal planning.25 The present generation may choose any path that provides a constant or
increasing level of welfare.
This approach runs against some of the same problems confronting social welfare analysis.
By what standards, for example, are we to assess the welfare of future generations? One practical approach
might be to define sustainability as nondecreasing per capita consumption. Under this standard, sustainable
paths will exist whenever constant consumption paths are technically feasible. But aggregate economic
indicators are notorious for their neglect of nonmarket environmental amenities and the degradation and
depletion of natural resource stocks. Application of this approach will thus at a minimum require a careful
reconsideration of conventional accounting techniques.
An alternative approach to the definition of sustainable development focuses on the
conditions required to support a high standard of living into the indefinite future rather than the distribution
of welfare across generations per se. Thus sustainability implies that we should ensure "the ability of future
GW4-38
-------
Session GW-4
Howarth
generations to meet their own needs"14 or that future generations inherit "an undiminished natural and
economic endowment."15 This approach does not require an exact definition of the welfare of future
generations. But it does imply an obligation to conserve environmental quality for the benefit of future
persons.
A second issue is rooted in the inherent uncertainty concerning the future course of
economic development. Policy makers are in fact choosing a probability distribution of potential outcomes,
not a single well-defined path for the economy. Thus the question of risk is fundamental to intergenerational
resource policy. How far are we willing to go to protect future generations against the possibility of an
inhospitable world? As is argued above, the composition of future generations will depend on the state of
the world prevailing when they are born. The individuals alive at a particular date under alternative
contingent states should thus be regarded as ethically distinct potential generations, and sustainability would
seem to require that the welfare of each potential generation be equal to or greater than that of its
predecessor. Thus, in a world of uncertainty, the sustainability criterion may require sacrifices on the part of
the present generation not only to raise the expected welfare of future generations but also to ensure that
living standards are nondecreasing even under the worst of circumstances.27
This is a strong supposition that needs to be placed in the context of competing social
values. Few would argue, for example, that 50% of world income should be diverted to the construction of a
planetary defense system to protect against the slight risk that future generations would be left destitute
following a collision between the Earth and a large asteroid. On the other hand, the world community has
decided to incur significant costs to reduce the uncertain threat posed by ozone depletion in the upper
atmosphere. At a bare minimum, the sustainability rule suggests the moral obligation to take steps to reduce
threats to future generations if so doing does not noticeably impact the subjective welfare of existing persons.
This rule, termed the precautionary principle,,28'29 calls for the general reduction of risks to future welfare,
and mandates above all that we provide future generations with the flexibility required to adapt to
unforeseen and unforeseeable events.
4.0 APPLICATIONS OF THE PRECAUTIONARY PRINCIPLE
To get a sense of the operational significance of the precautionary principle, it is helpful to
state the principle hi clear and explicit terms:
(P) Inhabitants of today's world are morally obligated to take steps to reduce catastrophic risks to
members of future generations if doing so would not noticeably diminish their own quality of
life.
Whether one accepts or rejects this principle involves a value judgement, and revised versions of the principle
may be proposed. Malnes,30 for example, argued that the present generation is obligated to "revoke risky
activities that jeopardize future needs for the sake of less urgent contemporary interests . . . well beyond the
minimum requirements of subsistence." This matter is a question of social ethics that is best resolved
through the political process mediated by the moral convictions of the participants rather than through
technical analysis. A case may be made, however, that P or some related principle is both morally plausible
and a reasonable reflection of prevailing social attitudes towards environmental risk.
Consider the application of this principle to the following sets of facts:
1. A certain pesticide reduces losses for a number of crop species, thus yielding small
improvements in farm profitability. While the immediate health risks of using the
chemical are small, cumulative use leads to irreversible ground water contamination
that scientists believe may cause serious birth defects and childhood cancer
fatalities. A substitute technology is available that would eliminate these risks yet
impose a small increase in the price of some fruits and vegetables.
GW4-39
-------
Session GW-4
Howarth
2. Chlorofluorocarbons (CFCs) released to the atmosphere deplete the stratospheric
ozone layer, increasing the proportion of ultraviolet-B radiation reaching the Earth's
surface. Because CFCs accumulate and persist in the atmosphere for nearly a
century, today's emissions will have a disproportionate impact on future generations.
Impacts on human welfare are uncertain but are thought to include increased
deaths from skin cancer and potentially serious damage to agricultural and natural
ecosystems. CFC substitutes are readily available but would impose modest cost
increases in certain products and processes: refrigeration systems, air conditioners,
aerosol sprays, and the manufacture of electronic goods.
3. Anthropogenic emissions of greenhouse gases such as carbon dioxide and CFCs
threaten to raise global temperatures and alter weather patterns in unpredictable
but potentially alarming implications for the welfare of future generations. Impacts
might include sea-level rise, storm intensification, increased frequency of droughts
and floods, reduced agricultural yields, mass species extinctions, disturbance of
natural ecosystems, and increased prevalence of tropical diseases. Stabilization of
current climatic conditions would impose large social costs. But steps to limit
cumulative warming to no more than 2° C could be taken without noticeably
reducing the subjective well-being of present or future persons.31'32
Under each set of facts, the requirements of the precautionary principle are clear: The
government should act to reduce risks to future generations by banning the use of hazardous chemicals that
yield trivial benefits, phasing out the production of CFCs, and imposing policies that reduce anthropogenic
emissions of carbon dioxide. Whether such recommendations would follow from the application of
conventional cost-benefit analysis is difficult to determine. In all likelihood we could not confidently quantify
the impacts of these environmental insults even in physical terms. Monetization would then constitute a leap
into the unknowable, rendering cost-benefit analysis inoperational for the cases under discussion.
In embracing the precautionary principle, we simplify the task of policy analysis and render
it operational by reducing it to a two-part test: Does a particular environmental insult impose catastrophic
risks on members of future generations? Can we take steps to reduce those risks without substantively
compromising our own well-being? Within this framework, there is ample room for the application of
economic and technical analysis. But objective analysis is a tool to be used in the identification and
characterization of policy impacts, not a substitute for the properly subjective elements of arriving at a
decision.
It is important to bear in mind that the "catastrophic risks" of the precautionary principle
are risks to particular persons and may or may not entail threats to the general integrity of social or
environmental systems. We might imagine a future world that is on the whole considerably richer than our
own. Yet it would be wrong for us to impose crippling burdens on some number of its inhabitants for the
pursuit of minor benefits to ourselves. A child born with a serious birth defect has no power over her
destiny, and it is difficult to conceive of remuneration sufficient to compensate her parents or herself for her
injuries. Nor can we claim that she would have been willing to accept the risk of deformity in exchange for
offsetting benefits under more favorable circumstances, for a person's moral identity is contingent on the
circumstances of her birth. The deformed person, born undeformed, would be an entirely different person
by genotype and sense of personhood.
Or suppose that our emissions of greenhouse gases would leave a particular peasant society
destitute by undermining the climatic conditions required to sustain fruitful agriculture. Even if the welfare
of our own descendants were not at risk, the precautionary principle holds that we would be obligated to
abate emissions, aid peasants to reduce their vulnerability to climate change, or both-provided that such
actions would not noticeably reduce our subjective well-being.
The strengths of the precautionary principle include its informational economy and explicit
foundations in normative values frequently articulated by participants in debates over environmental policy.
GW4-40
-------
Session GW-4
Howarth
As we have seen, the approach is often helpful in evaluating long-term environmental problems characterized
by substantial uncertainty. In itself, the precautionary principle is a partial guide to policy that is best
considered in the context of other planning criteria. In this sense, the approach challenges policy analysts to
adopt a strategy of methodological pluralism33'34 weaving together the insights gleaned from complementary
scientific, ethical, and economic frameworks to achieve a synthetic view that is greater than the sum of its
parts.
5.0 REFERENCES
1. Broecker, W.S. "Unpleasant Surprises in the Greenhouse." Nature. Vol. 328, pp. 123-126.
1987.
2. Smith, V.K. Environmental Policy under Reagan's Executive Order: The Role of Benefit-
Cost Analysis. Chapel Hill. University of North Carolina Press. 1984.
3. Johansson, P.O. The Economic Theory of Environmental Benefits. Cambridge.
Cambridge University Press. 1987.
4. Johnson, R.L. and G.V. Johnson. Economic Valuation of Natural Resources: Issues.
Theory, and Applications. Boulder. Westview Press. 1990.
5. Willig, R. "Consumer Surplus without Apology," American Economic Review. Vol. 66,
pp. 589-597. 1976.
6. Howarth, R.B. and R.B. Norgaard. "Environmental Valuation under Sustainable
Development." American Economic Review Papers and Proceedings. Vol. 82, pp. 473-477.
1992.
7. Dasgupta, P.S. and G.M. Heal. Economic Theory and Exhaustible Resources. Cambridge.
Cambridge University Press. 1979.
8. Daly, H.E. and J.B. Cobb. For the Common Good: Redirecting the Economy Toward
Community, the Environment, and a Sustainable Future. Boston. Beacon Press. 1989.
9. Howarth, R.B. "Intergenerational Justice and the Chain of Obligation." Environmental
Values. Vol. 1, pp. 133-140. 1992.
10. Wilson, R. "Risk Measurement of Public Projects," Discounting for Time and Risk in
Energy Policy. Resources for the Future, Washington, pp. 205-249. 1992.
11. Howarth, R.B. "Economic Efficiency, Intergenerational Equity, and Uncertainty: The
Theory of Climate Policy," Paper presented to the Peder Sather Symposium on Global
Climate Change, Berkeley, California. October 16-18, 1991.
12. Sen, A.K. Collective Choice and Social Welfare. San Francisco. Holden-Day. 1970.
13. Tietenberg, T. Environmental and Natural Resource Economics. Glenview, Illinois. Scott
Foresman. 1984.
14. World Commission on Environment and Development. Our Common Future. Oxford.
Oxford University Press. 1987.
15. Brown, L.R., C. Flavin, and S. Postel. "Picturing a Sustainable Society," State of the World
1990. New York. Norton, pp. 173-190. 1990.
GW4-41
-------
Session GW-4
Howarth
16. Schwartz, T. "Obligations to Posterity," Obligations to Future Generations. Philadelphia.
Temple University Press, pp. 3-13. 1978.
17. Parfit, D. "Energy Policy and the Further Future: The Identity Problem," Energy and the
Future. Totowa, New Jersey. Rowman and Littlefield. pp. 31-37. 1983.
18. Green, R.M. "Intergenerational Distributive Justice and Environmental Responsibility,"
Responsibilities to Future Generations. Buffalo. Prometheus, pp. 91-101. 1981.
19. Barry, B. "Intergenerational Justice hi Energy Policy," Energy and the Future. Totowa,
New Jersey. Rowman and Littlefield. pp. 15-30. 1983.
20. Dower, N. Ethics and Environmental Futures. International Journal of Environmental
Studies. Vol. 21, pp. 29-44. 1983.
21. Page, T. "Intergenerational Justice as Opportunity," Energy and the Future. Totowa, New
Jersey. Rowman and Littlefield. pp. 38-58. 1983.
22. Brown Weiss, E. In Fairness to Future Generations: International Law. Common
Patrimony, and Intergenerational Equity. Dobbs Ferry, New York. Transnational
Publishers. 1989.
23. Lele, S.M. "Sustainable Development: A Critical Review." World Development. Vol. 19,
pp. 607-621. 1991.
24. Pezzey, J. Economic Analysis of Sustainable Growth and Sustainable Development.
Washington. The World Bank. 1989.
25. Riley, J.G. "The Just Rate of Depletion of a Natural Resource," Journal of Environmental
Economics and Management. Vol. 7, pp. 291-307. 1980.
26. Repetto, R., W. Magrath, M. Wells, C. Beer, and F. Rossini. Wasting Assets: Natural
Resources in the National Income Accounts. Washington. World Resources Institute.
1989.
27. Howarth, R.B. "Intergenerational Competitive Equilibria under Technological Uncertainty
and an Exhaustible Resource Constraint." Journal of Environmental Economics and
Management. Vol. 21, pp. 225-243. 1991.
28. Perrings, C. "Reserved Rationality and the Precautionary Principle: Technological Change,
Tune and Uncertainty in Environmental Decision Making," Ecological Economics: The
Science and Management of Sustainabilitv. New York. Columbia University Press.
pp. 153-167. 1991.
29. Goodland, R. and G. Ledec. "Neoclassical-Economics and Principles of Sustainable
Development," Ecological Modeling. Vol. 38, pp. 19-46. 1987.
30. Malnes, R. The Environment and Duties to Future Generations-An Elaboration of
Sustainable Development. Lysaker, Norway. Fridtjof Nansen Institute. 1990.
31. Krause, F., W. Bach, and J. Koomey. Energy Policy in the Greenhouse. Volume One.
From Warming Fate to Warming Limit: Benchmarks for a Global Climate Convention. El
Cerrito, California. International Project for Sustainable Energy Paths. 1989.
GW4-42
-------
Session GW-4
Howarth
32. Howarth, R.B. and PA. Monahan. Economics. Ethics, and Climate Policy. Berkeley.
Lawrence Berkeley Laboratory. 1992.
33. Norgaard, R.B. "Environmental Economics: An Evolutionary Critique and a Plea for
Pluralism," Journal of Environmental Economics and Management. Vol. 12, pp. 382-394.
1985.
34. Norgaard, R.B. "The Case for Methodological Pluralism." Ecological Economics. Vol. 1,
pp. 37-57. 1989.
GW4-43
-------
Session GW-4
Questions and Answers
Questions and Answers
(transcribed from audio tape)
QUESTION: Do laws, regulations, or the courts recognize non-use values for ground
water? Either of you want to take a shot at that one?
MR. HARMON: I will try to take a stab at that. In my experience, the courts are
beginning to recognize such non-use values. The traditional beneficial uses that we have been talking about,
the consumptive types of uses, of course, are the traditional ones recognized in courts of law but hi my
experience in recent years, the past eight or ten at least, the courts that I had occasion to either visit or
testify in or see readings from, we are beginning to see decrees, water rights values being conferred upon
wildlife, fish propagation and this is natural that I am speaking of, in stream flows, wilderness values. We
just went through a tremendous fight in Colorado and it probably is not over yet, concerning the preservation
of wilderness values through minimum in-stream flows in headwaters areas of the state. So in my
experience, finally, again, they are beginning to be addressed and water rights are conferred in courts of law
for these non-use values.
DR. CRUTCHFIELD: I will amplify that a little bit. First of all, I should say I am not a
lawyer, I am an economist. I do not know which I should be.
MR. HARMON: Neither am I.
DR. CRUTCHFIELD: I do not know which I should be apologizing for. And again, I
should also remind people of my role. I am a federal government economist so I am not party to a lot of
the lawsuits and lot of litigation and the rather lucrative financial opportunities available with my colleagues
from universities to consult on these sorts of things.
But having said that, the classic case, of course, is the Exxon Valdez case and there was a
lot of work that was done in sort of contingent valuation, nonmarket estimates and cost estimation type
framework which is currently before the courts.
At the meeting of the American Agricultural Economics Association in Baltimore this
summer, there was a very interesting session where we talked about nonmarket valuation issues and a lot of
it had to do with non-use values and values that people place on seals and wildlife impacts and these sorts of
things that the participants could not discuss because they were under a court order to be quiet about it.
But there is, as I understand it, a substantial amount of regulatory and legal framework that
says that in valuing damages to natural resources from things like oil spills and other environmental mishaps,
that non-use values to have a role and that economics can offer something in the standpoint of, for example,
some of these contingent valuation studies.
I will say, however, that my one experience with this is when I was a professor at the
University of Rhode Island. I was a consultant on a case in the New Bedford Harbor where we were trying
to value the damages, the closure of the lobster fishery, and my component was to do some nonmarket
benefit estimates or cost estimates of the value of killing the lobster. The lawyers decided not to take the
case to court because they did not want to try to go before this particular judge and have an economics
lecture, so I was released from my role in that case, so yes, there is a role for nonmarket and non-use
benefits estimates but I think a lot of it depends on the particulars of the case.
QUESTION: Does concern for future generations require bequest of the same resources
or the capability of obtaining the same outcomes or the same levels of welfare?
GW4-44
-------
Session GW-4
Questions and Answers
DR. HOWARTH: That is a tricky question. If we played the game of neoclassical general
equilibrium models and we assumed perfect foresight in all of that, then the answer is no, it does not matter
what we transfer. If we play the game where we are not in a neoclassical role, we do not know what the
future might look like, we do not know the preferences or the needs of future generations, then it might be,
and I am putting this out as a conjecture, that we need to worry a lot about making sure that we preserve
base levels of certain kinds of assets. In a sense, this is the heart of what the debate about sustainability is
about. Sustainability, after all, can be addressed in an neoclassical general equilibrium model with perfect
foresight, simply by introducing intergenerational transfers and then this problem goes away.
Sustainability means choosing intergenerational transfers and then we use cost-benefit
analysis to get greater efficiency but figuring out what happens in the world that is complicated and in a
sense unknowable, I think that is an interesting question, but I think a lot of work is likely to go into that in
Dr. Costanza's society.
QUESTION: How do you approach the issue of discounting in the context of bequest
values?
DR. CRUTCHFIELD: Very carefully. In theory, if we had perfect information and we had
all the technical experts who could tell us the time path, for example, of rate of pollutant accumulation in an
aquifer, in our particular case, of ground water, or, for example, in the case of a wildlife or endangered
species where we may have a hit and that the time path of recovery. For example, that is the way some of
this is done with the oil spill models in terms of value resources.
In theory, if we have perfect information and knew the pattern over time at which the event
would either manifest itself in terms of growing pollution or the cleanup and then recovery of a natural
ecosystem to former values and then we had an appropriate choice of discount rate, then we could bring
everything back to the future, from the future to current values.
It is like everything else we do around here. We have to make best guesses at it. I think
there is a role for discounting, but it still means that we have to answer the questions of the ecological
impacts of the environmental science that we as economists have great difficulty handling. I do not want to
even jump into your question about whether or not discounting is ethical or moral or not. I would say there
is a role for it but it is a difficult issue.
DR. HOWARTH: (inaudible) is making an allocation more efficient, and to come back to
the former question, the issue here, really, is in a sense, cost-benefit analysis is about this balancing act of
how much environmental quality versus how much capital do we provide, and that is a separable question for
how much natural resources and capital do we provide. And the discount rate comes from the decision
about how much, what level of asset transfer do we settle on.
QUESTION: What approach should we take in valuing stressed ground water that takes a
long time to recover versus waters that can recover quickly? This has to do, I assume, with the vulnerability
of the resource, how sensitive it is, not that it is necessarily contaminated now but the question, what
approach do we take to value stressed ground water that takes a long time to recover versus waters that can
recover quickly?
SPEAKER: I have something to say about the program. That means to me ground water
is contaminated with a chemical based (inaudible).
DR. CROPPER: Well, it seems like a simple answer would be that if you have the stress
case where you have got the damages that in the absence of doing anything, are going to continue for a
longer time, the benefits of cleanup, you are really avoiding damages that would have gone into the future
for a longer time that, in the case where there is some kind of natural cleansing going on that enables the
aquifer to somehow cleanse itself. So it really affects the path of benefits, how long the benefits work.
GW4-45
-------
Session GW-4
Questions and Answers
DR. COSTANZA: You are also talking about the general issue of irreversibility and I also
think the uncertainty beyond that, but irreversibility as well, are there certain things that will damage the
environment that cannot be undone, and I think that gets into some of the ethical issues that Rich Howard
was talking about, and whether under conditions of irreversibility, we should use cost benefit analysis or
whether we should use something more akin to safe minimum standards, especially if we do not know what
the impacts of removing that environmental resource from society's views are going to be.
SPEAKER: This is a case where technology does not allow you to try, to do something.
You can make some minimal change perhaps but beyond that, the change is very small so you know the
damage is there.
DR. COSTANZA: There are irreversible effects that may be trivial in terms of their real
value to society. There are other irreversible effects that may be extremely important, and so you still need
to have some relative scale for a resisting the trivial or important effects, but I think it takes you outside of
the traditional framework, which is what we are talking about, irreversible effects that are important.
DR. HOWARTH: It depends what you see as the potential consequences and then you
want to go back and you want to say, what is at stake and why do we care about this problem? It is hard to
give a generic answer to your question without looking at the context of what is happening with the
examples.
DR. CROPPER: But also, I think the importance of irreversibility depends on how much
you think, which was an earlier question, how much you think ground water is substitutable for other goods.
People who think you have to get the clean ground water to future generations, for those people, the
irreversibility has very important consequences that it does not have and you can substitute other goods or
clean ground water.
QUESTION: The question is, what ways can economists and ecologists work better to
address these problems?
DR. HOLLAND: Actually, I was thinking about the way that we each come at
sustainability, for one. The folks who are involved in implementation of the sustainable bias here, initiative,
have been working now with economists and geographers and ecologists and geologists to try to come up
with a common definition for what we are all dealing with here and, in fact, each of us comes with our own
disciplinary baggage.
I think one of the most important things that all of us need to deal with is to try to talk
across disciplinary barriers and I think that is, to the ecological society, what the whole SBI is all about, is to
try to work with other disciplines and to try to come to grips with some of this disciplinary baggage that each
of us has come up with, and see if we cannot come up with some common terminology so we are all
speaking the same language.
DR. COSTANZA: Ours is a society with over 1,300 members now devoted to exactly that
issue. We tried to break out those disciplinary barriers and get some effective dialogue going.
MR. JOB: Okay, I know it is late in the day and I do not want to keep you a whole lot
longer but I would like to make a few observations real quickly here and then we will call the session to a
close. In reflecting on the entire discussion of this panel and earlier panels, I was making some notes on
words and phrases that kept coming up both in an environmental or ecological context and an economics
context and it is sort of an interesting patchwork of words and phrases, things like interconnections. The
economy is interconnected. The environment is interconnected. Transfers. Interdependence. Both are very
interdependent within themselves and between each other. Long-term sustainability, economics and
environmental. Direct and indirect impacts, obviously things we are concerned about we try to measure.
Uncertainty, the range of possible outcomes, both economic outcomes, environmental and ecological
outcomes.
GW4-46
-------
Session GW-4
Questions and Answers
These are all things that sort of weave the two areas together. I think that there are things
that frame considerations that EPA needs to take into the future in its own regulatory programs, and I just
want to mention in closing that our comprehensive ground water protection program is concerned with these
kinds of considerations.
We are working with the states, recognizing their key role in protecting ground water
resources and asking them to consider the value of the resource, the value of the ground water resource, and
we are asking them to consider how these factors, the factors that have been discussed today will cause them
to value their resource, set policies for that resource, set priorities for future action.
If any of you are not aware of our comprehensive state ground water protection program
initiative, we are preparing a guidance document that the states will be using in developing that program.
You are certainly welcome to receive a copy of our guidance document. We will be putting that out in final
form in about a month, and I would invite you to participate in that process in the state you are a resident
in, to guide the development of these programs and help reflect the values that you think we should be
reflecting in these programs for the ground water resources.
GW4-47
------- |