I
I
.': '••Afjril:l983/..
I
i
I
.w
Benefit-Cost Assessment
. U.S. EPA Headquarters Library
•;•''"'. . Mail code .3201.' '•"'••;
-' • ^ 1200'Pennsylvania Avenue NW
s.;" : Washington DC 20460; .
Programs
Volume I
\A
f
I
-------
I
I
I
I
if*0
April 1983
I
0-
I
I
I
I
I
I
I
I
I
I
I
I
Benefit-Cost Assessment Handbook
for Water Programs
Volume I
Prepared for
U.S. Environmental Protection Agency
Economic Analysis Division
Washington, DC 20460
U.S. EPA Headquarters Library
Mail code 3201
1200 Pennsylvania Avenue NW
Washington DC 20460
Prepared by
Dr. William H. Desvousges
Research Triangle Institute
Research Triangle Park, NC 27709
and
Dr. V. Kerry Smith
University of North Carolina
Chapel Hill, NC 27514
The information in this document has been subject to the U.S. Environmental
Protection Agency's (EPA) Peer and Administrative Review, and it has been ap-
proved for publication as an EPA document. Mention of trade names or com-
mercial products does not constitute endorsement or recommendation for use.
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
i
i
April 1983
Benefit-Cost Assessment Handbook
for Water Programs
Volume I
Prepared for
U.S. Environmental Protection Asency
• Economic Analysis Division
Washington, DC 20460
Prepared by
Dr. William H. Desvousses
Research Triansle Institute
Research Triansle Park, NC 27709
and
Dr. V. Kerry Smith
University of North Carolina
Chapel Hill, NC 27514
The information in this document has been subject to the U.S. Environmental
Protection Agency's (EPA) Peer and Administrative Review, and it has been ap-
proved for publication as an EPA document. Mention of trade names or com-
mercial products does not constitute endorsement or recommendation for use.
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
PREFACE
As one component of the decision process, water program offices at the
Federal, State, and local levels along with advisory citizen groups have per-
ceived the need to consider the beneficial and detrimental effects of policies
that affect water quality. In response, the U.S. Environmental Protection
Agency's Office of Policy Analysis sponsored the development of this Benefit-
Cost Assessment Handbook for Water Programs. This experimental handbook
will not short-circuit any Federal, State, or local water quality laws. Rather,
It provides suggestions about how to evaluate the economic aspects of a pro-
posed policy as a regular part of the decision process. These evaluations can
identify water quality policies that have highest priority, so that society's
resources can be directed to the areas that will have the greatest benefits.
They also can serve as one of several analyses that support the decision proc-
ess for any specific water quality policy. Such ah approach can help assure
the attainment of our Nation's water quality goals with a minimum expenditure
of resources.
The Benefit-Cost Assessment Handbook for Water Programs is a primer.
It assumes only a limited familiarity with economics. Throughout the handbook,
case studies help to clarify points. Data needs, key assumptions, and other
relevant points are covered for different ways of determining the relationship
between desirable and undesirable effects of a program decision.
Since water program offices have begun to move toward the use of
benefit-cost concepts, the scope of the handbook is broad enough to explain
how to conduct benefit-cost assessments in diverse applications. The costs
and health benefits of drinking water policies are covered elsewhere, so this
handbook concentrates on the benefits and costs for other water programs.
Although most of the examples in this volume are for hypothetical water quality
standards decisions, the tools can be applied to a broad spectrum of water
quality decisions, and even to environmental issues in general.
If there is sufficient interest in this experimental approach, a second
volume may be developed to provide more in-depth discussion of the benefit-
cost assessment for use by practitioners. It also may be desirable to tailor a
similar document for specific water programs. The Office of Policy Analysis
welcomes comments and suggestions, which may be directed to:
Ann Fisher
U.S. Environmental Protection Agency
401 M Street, S.W.
Office of Policy Analysis (PM-220)
Washington, D.C. 20460
(202) 382-2783
iii
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
ACKNOWLEDGMENTS
Why are the acknowledgments always the last section to be written
in any report? This question, which often .occurs on the way to work
toward the end of a project, is likely one without a definite answer.
We could surmise that it comes from an unwillingness to let go at the
end of any project (not likely) or simply that we have saved the best
for last.
"i
We are indebted to many people who have given of their time to
assist us in writing this handbook. Yet there is one person whose
commitment to the project stands out. That person is our project
officer at the Environmental Protection Agency (EPA), Dr. Ann Fisher.
The simple truth is that without Ann Fisher this handbook would not
have been written. With the many hours she has given to the project
she is more author than anything else. Since she would never have
allowed this acknowledgment to appear .in the report had she read it,
perhaps this is the reason for leaving it until last.
As this handbook has evolved, it has drawn from many sources.
Chapter 4 pertaining to the costs of regulatory actions draws heavily
on an earlier chapter prepared by Metasystems, Inc., and particularly
Tze-Wen Chi and Peter Morgan. Indeed, most of the examples are
taken from this earlier draft. Assistance on the cost chapter came
from the Office of Analysis and Evaluation under Louis Dupuis. John
Kukulka and Joe Yance from the staff suggested many helpful revisions
in support of the project.
Office of Policy Analysis staff members provided valuable detailed
critiques of early drafts. Reed Johnson, Bob Raucher, Peter Caulkins,
and Skip Luken gave good counsel. Joan O'Callaghan's valuable edi-
torial suggestions on an early draft have been incorporated into this
draft.
The EPA staff in the water program area have also played an
important role in this project. Patrick Tobin, Dave Sabock, xand
Marjorie Pitts of the Office of Water Regulations and Standards, Myron
Temins and Charles Moar of the Office of Water Programs Operations
(OWPO), and Jerry Manorola, formerly of OWPO, gave extensive
comments.
In addition, helpful comments were received from Dan Huppert,
MNFS Southwest Fisheries Service; Robin Gottfried, University of the
South; John Luensman, Department of Planning and Development,
Chautauqua County, New York; John Loomis, U.S. Forest Service;
and Mike Piette, University, of Hartford.
v
-------
Finally, the RTI staff has been most supportive. Tayler Bingham
brought his experienced view of environmental regulations to bear on
all of the drafts of this handbook. Tayler's insightful comments are
valued , highly. Hall Ashmore's editorial efforts mark every page of
this handbook. Hall's economy with words and his ability to make an
author's prose bear fruit are assets for any handbook or report.
Finally, Jan Shirley and the Institute's Word Processing Specialists
have done yet another outstanding job of producing a pleasing report.
Jan's devotion to quality and her ability to know when to coax or
cajole a reluctant writer are most appreciated.
VI
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
CONTENTS
Chapter Page
1 BENEFIT-COST ASSESSMENT: A COMMON
SENSE APPROACH TO DECISIONS 1-1
1.1 introduction 1-1
1.2 Regulation: An Overview 1-2
1.3 Benefits: An Economic View 1-4
1.4 Costs: An Economic View 1-5
1.5 Benefit-Cost Assessement: What Is It? 1-6
.Is Benefit-Cost Assessment Different
from Cost-Benefit Analysis? 1-6
Benefit-Cost Assessment: ' A Step-by-
Step View 1-7
1.6 Key Considerations in an Assessment 1-11
1.7 Summary 1-12
1.8 Guide to Handbook 1-13
2 ISSUES IN A BENEFIT-COST ASSESSMENT 2-1
2.1 Introduction 2-1
2.2 Including Intangibles in a Benefit-Cost
Assessment 2-1
Introduction 2-1
. Example 2-2
2.3 Economic Impact Measures 2-5
Assessing Household Impacts 2-5
.Ability to Pay 2-5
Financial Capability 2-6
• Assessing Industrial Impacts 2-6
Assessing Changes in Employment,
Output, and Prices 2-6
2.4 What to Do About Distribution: Problems
in Adding up Over People 2-7
VII
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
CONTENTS (continued)
Chapter
Introduction 2-7
Example I 2-8
Example II. 2-9
Summary: Distribution 2-9
2.5 Discounting Future Benefits and Costs:
Adding up Over Time . 2-10
Time Preference: What Is It? 2-11
How To Determine the Discount Rate in
a Less Than Ideal World 2-12
Including Risk in an Assessment of
Discount Rates 2-14
What Are the Empirical Implications of
the Discount Rate Issues? 2-15
The Simple Mechanics of Discounting . 2-16
Discounting: A Summary Review 2-18
2.6 Summary 2-19
3 MEASURING THE BENEFITS OF WATER
QUALITY PROGRAMS 3-1
3.1 .Introduction 3-1
3,2 Categories of Benefits: Ah
Overview 3-1
3.3 Practical Concerns: Selecting a Benefit
Estimation Method 3-4
3.4 Household Benefits : 3-5
Theory: The Demand Function " 3-5
Practice: Methods for Measuring
Household Benefits • 3-6
Travel Cost Method .' 3-6
Data Needs and Key Assumptions and
Features 3-8
Case Study: Alternative Approaches for
Estimating Recreation and Related Benefits
of the Monongahela River . 3-9
viii
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
CONTENTS (continued)
Chapter
Survey-Contingent Valuation -. . . . 3-14
Data Needs, Key Assumptions/
Limitations, and Features 3-16
Case Study: Recreation and Related Benefits
of Water Quality Improvements in the
Mondngahela River 3-18
Survey—Recreation Participation Models 3-20
Data Needs, Key Assumptions/
Limitations, and Features 3-21
Case Study: A Participation Survey
Approach to Valuing Water Quality
Improvements 3-22
Hedonic Property Value Method 3-25
Data Needs, Key Assumptions/
Limitations, and Features 3-26
. Damage Function Method 3-27
Data Needs, Key Assumptions/
Limitations, and Features 3-27
3.5 Business Benefits 3-27
Theory: The Supply Function 3-27
Practice: Cost Savings Function Method for
Measuring Business Benefits 3-28
Data Needs, Key Assumptions/
Limitations, and Features . 3-29
Case Study: Irrigation Benefits . . . 3-30
3.6 Public Water Supply Benefits 3-32
3.7 Summary 3-32
4 MEASURING THE COSTS OF WATER QUALITY
PROGRAMS 4-1
4.1 Introduction 4-1
4.2 Measuring Costs: The Basic Concepts 4-1
4.3 Measuring Costs: Two General Approaches ....... 4-3
4.4 Types of Costs 4-4
Capital Costs '. 4-4
IX
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
.CONTENTS (continued)
Chapter
Operation and Maintenance Costs 4-5
Reinvestment Costs 4-6
Salvage Value '. 4-6
Summary Cost Measures 4-8
4.5 Practical Issues in Measuring Costs 4-8
Sources of Cost Data 4-8
The Use of Cost Indexes 4-9
Major Factors Affecting Cost Estimates 4-11
Sample Data Form 4-12
4.6 Cost Estimating Techniques 4-12
Components of Treatment Systems 4-12
Estimating treatment Costs 4-15
4.7 Examples Using Cost Techniques 4-17
Example 1: Pulp Mill Using EPA Development
Document 4-17
Example 2: POTW Using CAPDET 4-21
Example 3: POTW Using Technology
Assessment Manual . 4-25
Example 4: What to Do When There Is
"No Information" 4-25
4.8 Final Cautions 4-27
4.9 Summary 4-27
5 COMPLETING THE BENEFIT-COST ASSESSMENT 5-1
5.1 Introduction 5-1
5.2 Sensitivity Analysis: A Gauge to Believability .... 5-1
Introduction 5-1
Example 5-3
Step 1: Translate the Benefits and
Costs into Present Values 5-3
Step 2: Perform Sensitivity Analysis for
Discount Rate and Key Assumptions 5-4
Step 3: Interpret Sensitivity Analysis 5-4
-------
CONTENTS (continued)
Chapter Page
5.3 Displaying the Assessment Results 5-4
Narratives 5-6
Arrays 5-6
Graphs •. 5-6
5.4 Benefit-Cost Assessment Checklist . . . .- 5-9
5.5 Summary 5-9
6 BENEFIT-COST--SAMPLE SCENARIOS 6-1
6.1 Introduction 6-1
6.2 Simple Case Scenario 6-2
Introduction 6-2
Simple Case Scenario Format . 6-2
6.3 Medium Case Scenario 6-4
.Introduction 6-4
Medium Case Scenario Format. . . . ' 6-5
6.4 Complex Case Scenario 6-6
Introduction 6-6
Complex Case Scenario Format .' . . 6-7
6.5 Summary 6-26
7 References 7-1
XI
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
FIGURES
Number
1-1 Effects and responses to water quality
regulatory actions 1-2
1-2 Key steps in a benefit-cost assessment .'.... 1-9
1-3 Key steps in a complex benefit-cost
assessment . 1-10
3-1 A spectrum of water quality benefits 3-2
3-2 The demand function and consumer surplus 3-5
3-3 Illustration of the treatment of water quality
with travel cost demand model 3-7
3-4 General travel cost demand model for a water
quality improvement 3-13
3-5 Supply .function and the producer surplus 3-28
3-6 Irrigation benefits ' 3-30
4-1 Measurement of opportunity/costs 4-2
4-2 Sample model plant data form 4-13
5-1 Tradeoff curve. 5-7
5-2 Checklist for a water quality standards
benefit-cost assessment 5-8
6-1 Travel cost demand function and consumer
surplus with boatable water 6-10
6-2 Travel cost demand function and change in
consumer surplus with fishable water 6-12
6-3 Travel cost demand function and change in
consumer surplus with swimmable water . . 6-16
XIII
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Number
2-1
2-2
2-3
2-4
2-5
2-6
TABLES
Benefits and Costs of Attaining Fish and
Wildlife Use: Array 1 2-3
Benefits and Costs of Attaining Fish and
Wildlife Use: Array 2 . 2-3
Benefits and Costs of Attaining Fish and
Wildlife Use: Array 3 2-4
Summary of Industrial Impact Measures 2-7
Distribution of Benefits and Costs 2-9
Summary of Final Description Factors
Influencing Shadow Price of Capital 2-13
3-1 Classification of Benefit Categories 3-3
3-2 Site Attributes Considered in Monongahela
River Study 3-11
3-3 Estimated Equations for Site Demand Parameter
Estimates
3-11
3-4 Specifications for the Dissolved Oxygen
(DO) Levels Associated With Use Designations 3-12
3-5 Benefit Estimates from Generalized Travel
Cost Model With the Monongahela Survey
Respondents 3-14
3-6 Summary of Biases in Contingent Valuation
Experiments 3-17
3-7 Willingness to Pay for Three Levels of
Water Quality 3-19
3-8 Vaughan-Russell Model—Predicted Effects
of BPT Regulations on Participation . 3-24
4-1 Components of Capital and .Operation and
Maintenance Costs 4-5
4-2 Variables and Definitions for Measuring Costs 4-7
4-3 Example 1: Cost Summary-Market Bleached
Kraft, Subcategory 4-18
xv
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
TABLES (continued)
Number
4-4 Example 2: Current and Upgraded Treatment
Trains for CAPDET . 4-22
4-5 Example 2: Price and Cost Inputs to CAPDET 4-23
4-6 Example 2: POTW Upgrading Cost Summary 4-24
4-7 Example .3: Development of Capital Costs
(Nitrification) 4-26
4-8 Example 3: Summary of Costs for Nitrification
Upgrade 4-26
5-1 Key Elements of Benefits and Costs . .' 5-2
5-2 Sensitivity Analysis Calculations: Discount
Rate at 4 percent . . 5-5
5-3 Sensitivity Analysis 5-5
6-1 Demand for Recreation for River 1--Water
Quality at Level Suitable for Boating 6-9
6-2 Demand for Recreation for River 1--Water .
Quality at Level Where Gamefish (Bass) Can
Live in River 6-11
6-3 Capital Equipment and Cost for Model Plant
to Meet Regulation • 6-17
6-4 Estimated Capital Costs 6-18
6-5 Cost of Treatment Plant, City B 6-18
6-6 Present Value of Benefits . 6-22
6-7 Present Value of Operating Costs 6-23
6-8 Total Project Costs ' 6-23
6-9 Benefits and Costs of Attaining Fish and
Wildlife Propagation Use Designation 6-24
6-10 Distribution of Benefits and Costs 6-25
XVJ
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
CHAPTER 1
BENEFIT-COST ASSESSMENT: A COMMON
SENSE APPROACH TO DECISIONS
1.1 INTRODUCTION
Should a State change the designation of impaired uses for a river? Will
a sewer overflow project .provide benefits in excess of costs? Is advanced
waste treatment necessary to attain a river's designated uses? Are treatment
plants more desirable on some rivers.than on others?
This handbook—a primer on benefit-cost assessment—shows how economic
principles can help decisionmakers make these difficult choices. Its primary
intent is to demonstrate the common sense inherent in benefit-cost assess-
ments of alternative choices. In addition, this handbook shows how to add an
economic dimension to scientific and technical analyses in considering the full
range of impacts from a proposed water quality action.
Since they focus on the alternatives available to society, economic prin-
ciples are especially relevant to water quality program problems. Specifically,
the economic viewpoint recognizes that the scarcity of society's resources
forces choices among alternatives. However, choosing among alternatives
creates tradeoffs—i.e., one thing must be given up to attain another. Thus,
water quality decisions produce both desired and undesired effects for
society. Benefit-cost assessment simply uses economic principles to help the
decisionmaker make these choices.
Water quality programs implement regulatory mandates or provide assist-
ance either to those adversely affected by the regulations or for specific
projects. Principles covered in this handbook could be applied to:
Effluent guidelines issues that require limits on specific indus-
trial discharges.
Water quality standards issues where States designate uses for
water bodies and develop criteria to achieve the uses.
Advanced treatment issues where the Federal Government pro-
vides financial assistance to construct municipal treatment
plants that require advanced technologies.
Combined-sewer overflow issues where Federal assistance is
provided to deal with municipal runoffs that create pollution
problems.
1-1
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
This chapter discusses the basic principles in a benefit-cost assessment.
Specifically, Section 1.2 provides an overview of regulation, using linkages
between regulatory actions, effects, and changes in behavior. To highlight
the basic principles, Section 1.3 describes .benefits from an .economic perspec-
tive, and Section 1.4 details a similar discussion for costs. The basic con-
cepts of benefit-cost assessment are described in Section 1.5, along with a
step-by-step view of an assessment. Section 1.6 gives some groundrules for
performing an assessment and Section 1.7 summarizes the key points in the
chapter. Finally, Section 1.8 provides a guide to the remainder of Volume I
of the handbook.
1.2 REGULATION: AN OVERVIEW
Understanding how benefit-cost assessments can be used is easier with
some knowledge of how a regulation affects economic activities. The key to
this understanding is the linkage (shown in Figure 1-1) between (1) a change
in a regulation (an action), (2) its technical effects, and (3.) the behavioral
responses to it.
Water Quality
Regulatory Action(s)"
Change Designated Use(s)
Modify Criteria to Provide
for Designated Use(s)
Technical Effects
of Water Quality —-
Rcgulatory Action(s)
Changes in Effluents
Changes in Water Quality
Change in Ecological
Habitat
Effect* on Economic
Agents
Behavioral Effects I
of Water Quality _j
Regulatory Action(s) |
Behavioral Responses
of Economic Agents
Figure 1-1. Effects and responses to water quality regulatory actions.
1-2
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
One example of an action represented by the first two blocks in Figure
1-1 is a change in the uses designated for a water body and the associated
modifications of technical water quality criteria to accommodate these uses.
The action changes effluent levels and the resulting water quality and ecolog-
ical habitat—all of which affect households and businesses, the primary eco-
nomic agents. A change in effluent levels simply means more or fewer pollut-
ants will be discharged into the water body, thus altering overall water qual-
ity. The changes in water quality alter the diversity of microorganisms, fish,
or flora and fauna and can noticeably change the local ecological habitat. The
magnitude of the technical effects depends on specific water body characteris-
tics, the nature of the pollutant being controlled, and the extent of control.
For example, river depth, flowrate, and riverbed geology will influence the
technical effects of changing the designated uses of a river to include a warm-
water fishery.
Not all water quality programs are regulatory programs. For example,
the combined sewer overflow (CSO) program aims at directly enhancing water
quality by reducing the surge of pollutants following a severe storm. Even
in these programs, a determination of the linkages between the project and its
technical effects is essential.
Equally important to benefit-cost assessment is how businesses and house-
holds are affected by the action. For example, if improved water quality will
support a warmwater fishery in a watercourse, more gamefish will likely inhabit
the river, thus enabling fishermen 'to catch more fish--the technical effect on
the household. However, to achieve the level of dissolved oxygen necessary
to support gamefish, regulation might be required so that firms clean up their
discharges into the river. From society's viewpoint, therefore, actions have '
both beneficial and detrimental effects. Maximizing the public good requires
consideration of both types of effects.
Determining how beneficial and detrimental effects balance out requires
consideration of the final linkage—how primary economic agents change their
behavior in response to technical effects. For example, if the technical ef-
fects of a water regulation (e.g., an increase in gamefish populations) allow
fishermen currently using a watercourse to use it more, new users may be
attracted to the site. In economic terms, this situation is described as in-
creased demand for a site's recreation services. The amount of the demand
increase will be determined both by site attributes (features) and by the site
users. Important site attributes include the proximity of substitute fishing
streams, the number of access points, and the quality of local natural fea-
tures, such as the surrounding countryside. The incomes of the fishermen,
the price of fishing equipment, and how badly the users and potential users
want to fish—their preferences for fishing—will also affect the ultimate behav-
ioral response to increased gamefish populations.
However, just as households (or fishermen, as in the example) respond
to the technical effects of regulatory actions, firms also respond. Specif-
ically, they may decide to close down operations, alter waste treatment proc-
esses, or alter product mixes to meet the technical standard required by the
decision. Clearly, each of these behavioral responses has different conse-
1-3
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
quences, but the magnitude of the regulation's technical effects is determined
by the range of feasible responses and the market conditions for the goods
produced by the affected businesses. Thus, firms facing more favorable
market conditions find a wider range of choices open to them, and those
facing strong pressures from competing firms are more limited.
The actual regulatory process is considerably more complex and less cer-
tain than indicated above. For example, businesses using water in their
production process could be adversely affected by the regulation if higher
dissolved oxygen levels corrode their water pipes, resulting in higher operat-
ing costs. Similarly, individuals who are not users of the site may be af-
fected if they view general increased ecological diversity as a beneficial ef-
fect. In the end, therefore, practitioners will have to determine which ef-
fects are relevant for inclusion in the benefit-cost assessment.
More importantly, the linkages discussed in this section do not attach
values to the positive and negative effects. Rather, they merely suggest a
way of viewing the regulatory process to help determine what the effects are.
In some cases, however, the decision process is eased if effects are converted
to values. Unfortunately, the attachment of values to the effects is a trouble-
some process for many potential users of benefit-cost assessment.
1.3 BENEFITS: AN ECONOMIC VIEW '
The economic approach to defining and measuring regulatory benefits is
unfamiliar to many noneconomists. Quite simply, however, economics ap-
proaches benefits from society's perspective, assigning values based on indi-
viduals' willingness to pay for particular regulatory effects.* In essence,
economics implicitly assumes individuals are best suited to value the effects of
water quality programs.
Once both the beneficial and detrimental effects of a proposed action have
been identified, the practitioner may heed to weigh their relative importance
before a final decision is made. Of course, the economic valuing process
described above can help determine relative importance--e.g., area fishermen's
willingness to pay for an action to increase gamefish populations vs. the costs
incurred by a local industrial plant whose discharges the action will require be
cleaned up—but it has limits. Indeed, no approach—economic or otherwise—is
a substitute for the judgment that decisionmakers must exercise to make
choices among alternatives representing various types and degrees of well-
being to a variety of population subgroups (fishermen, plant owners, etc.).
Benefit-cost assessment is a framework for identifying and organizing informa-
tion to ease the decisionmaking process, not a decision rule.
*lt should be recognized, however, that, added up over all persons,
individual willingness to pay is influenced by the income, or wealth, available
to each person.
1-4
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
One frequently asked question is, "Why do people have to 'pay' for the
beneficial effects of water quality programs?" The answer is that they do not;
alternatively, they could accept payment to forgo the effects. In effect, the
two measures—willingness to pay and willingness to accept—are equally good,
but different, measures. While the "accept" measure implicitly assumes the
individual "owns" the rights to the beneficial change, the "pay" measure
assumes the opposite. Volume II of the handbook will discuss how benefits
based on these two different measures can be related. The important distinc-
tion is the equity question--!.e., whether individuals own rights to the pro-
gram benefits.
Although willingness to accept is an equally good measure, willingness to
pay is normally used to discuss regulatory benefits because it can be revealed
in markets, when they exist, through purchases of good's/or services affected
by the program. Then, benefits can be measured empirically. Although these
markets clearly do not exist for the effects of water quality improvements,
willingness to pay is still a useful way of valuing benefits. An assessment
may describe benefits only in qualitative terms, but the description can be
written from the perspective of willingness to pay. The benefits measurement
approaches discussed in Chapter 3 are ways that economists have approached
the benefits problem when markets do not exist. None of these approaches
gives precise estimates of willingness to pay. Each is a blunt tool, capable
only of giving rough estimates, which are sufficient in most cases.
1.4 COSTS: AN ECONOMIC VIEW
Opportunity cost measures the cost of any resource--e.g., labor, ma-
chinery, environmental resources—in terms of its next best alternative use.
That is, the value of forgone alternative uses of any resource provides the
basis for estimating the cost of any specific use. As a result, opportunity
cost considers tradeoffs—i.e., how much must be given up of one thing to
have more of another.
For example, assume a proposed project would improve a lake's water
quality to permit recreational fishing, boating, and swimming. The lake can-
hot now support any of these activities, but it would if quality were improved
by constructing a waste treatment plant along a river that feeds the lake. In
this example, the opportunity costs of the action would be the forgone op-
portunities of all the resources used in improving water quality. In the ab-
sence of market imperfections, the opportunity cost of construction inputs—
equipment, materials, labor, land, etc.--would be valued by their market
prices. In addition, if the action precludes use of the river or the lake for
other activities (such as industrial or agricultural uses), the values of these
forgone alternatives would also be part of the opportunity costs.
Many practitioners consider cost estimation an easier task than benefit
estimation. Perhaps a more accurate view is that many find it less objection-
able to value the labor, materials, and equipment used as a result of an
action. However, difficulties can arise when the full social costs of the in-
vestment alternatives are considered, or when effects on rates of technological
change are included. Cost estimation is likely to involve as many judgments
1-5
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
as benefit estimation and is subject to the same general cautions. Indeed,
caution is advised in making too great a case for precision in the measurement
of economic well-being when less precision than desired exists in the linkage
between the regulatory action and its effect on economic activities. The prob-
lems in establishing linkages do not imply these technical issues should .be
ignored, only that they be considered in their proper perspective—as a part
of the problem of evaluating the benefits and costs associated with a change
in water quality.
1,5 BENEFIT-COST ASSESSMENT: WHAT IS IT?
Benefit-cost assessment is a way of organizing information—a method for
identifying all the favorable and unfavorable outcomes of a proposed action.
Where necessary for complex decisions, many of these outcomes can be con-
verted into a common set of units (usually dollars) to permit consistent com-
parisions of benefits and costs. Monetization may be impossible for outcomes
which defy measurement. Even in these cases, the benefit-cost assessment
framework can organize information associated with an action. In short,
benefit-cost assessment is a practical method for including basic economic
principles in the decision process.
Although benefit-cost assessment is a guide to decisionmakers, it does
not provide the final answer to a public policy decision. Other factors such
as the public's view of appropriate uses for a particular water body or the
fairness of cost impacts on particular groups are important considerations.
What a .benefit-cost assessment does do is provide an organizing framework
for information the public and rulemaking body can use in making more in-
formed decisions.
It is important to recognize that value judgments are a part of all deci-
sions. Benefit-cost assessments supplement scientific and technical informa-
tion with economic information that may help decisionmakers make these judg-
ments. Very simply, a well-structured benefit-cost assessment can reduce
the complexity of what needs to be considered, making the decision process
more manageable.
Is Benefit-Cost Assessment Different from Cost-Benefit Analysis?
One of the first things that comes to mind for potential practitioners of
benefit-cost assessment is the past misuse of cost-benefit analysis. These
misuses emphasized the search for a ratio—the one "number"--that would ra-
tionalize or justify a project. In many instances, the misuses involved an
attempt to include benefits that were, at best, marginally related to a project.
This is not the case for the benefit-cost assessment suggested in this hand-
book. Since benefit-cost assessment requires a consistent, systematic treat-
ment of benefits and costs, an outside observer can easily discover when a
practitioner tries to stretch the approach beyond the limits dictated by com-
mon sense. When the assessment process is carefully conducted, common
sense will provide a reasonable guide through most of the decisions. Any-
thing not sensible should be scrutinized. Critics who maintain that benefit-
cost assessment can be used to justify anything overlook the fact that any
approach can be abused.
1-6
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Since it can compare benefits and costs in qualitative terms, in qualita-
tive terms with some quantification, in quantitative terms, or in monetized
terms, benefit-cost assessment is a more flexible approach than conventional
cost-benefit analysis. The key, of course, is the nature of the decision. A
qualitative assessment will reveal whether the potential benefits and costs at
stake in a water quality decision are clearcut. No further economic assess-
ment will be needed, yet the decisionmaker will have a logical, consistent
basis for economic considerations. If the situation is more complicated, or if
the potential benefits and costs at stake are considerably larger, a more de-
tailed benefit-cost assessment can make the decision more manageable.
Many ways exist to tailor a benefit-cost assessment to fit the needs of
the Issue at hand. For example, monetization can play a critical role in more
complicated benefit-cost assessment decisions. Specifically, by blending mone-
tization with qualitative judgment, .benefit-cost - assessment can determine
whether the benefits bear a reasonable relationship to the costs involved and
whether there will be significant impacts on certain parts of the population or
the economy.
A misconception that arises with benefit-cost assessment involves recent
techniques—such as survey techniques—developed to deal with previously
immeasurable or nonmonetizable benefits, such as enhanced ecological diver-
sity or amenities. These survey techniques are not opinion polls; they rely
on carefully designed questionnaires to measure an individual's willingness to
pay for these benefits. One fact has been clearly shown by all the survey
studies: such previously nonquantifiable benefits are indeed a substantial
component of the total environmental benefits picture. That is, the studies
have shown these benefits to be large, and an assessment that overlooks them
may indeed understate the full benefits. While most water quality decisions
will not require a survey to determine these previously nonquantifiable bene-
fits, some benefit-cost assessments may adapt the results of recent surveys
for specific sites (e.g., see Chapter 6, Section 6.3). A few instances may
occur in which, because the potential costs are so large, practitioners may
want to use simple surveys to get at least a ball park estimate of such poten-
tial benefits.
Benefit-Cost Assessment: A Step-by-Step View
Each of us makes decisions every day, judging whether the anticipated
consequences of an action will be "worth" the "costs." Of course, the mean-
ings of "worth" and "costs" vary from one person to the next because differ-
ent people evaluate the same action differently. Whatever the outcome, how-
ever, the logic underlying the decision process is the same. Based on this
decision logic, benefit-cost assessment is a method for defining "worth" and
"costs," offering a logical framework for structuring information for decisions
in the public sector.
Although performing a benefit-cost assessment is not a mechanical task
with each step completely known in advance, it is possible to outline the gen-
eral steps that are useful in assembling a complete assessment. These steps
1-7
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
flesh out the linkages between a policy decision and the behavioral changes,
highlighted earlier in Figure 1-1. For instance, an assessment can portray
the individuals and firms that will be directly affected by a regulatory action,
how they will be affected, and how they will change their behavior in response
to the regulation.
The logic of a benefit-cost assessment is straightforward, as shown in
the following steps:
Define the action
Determine an appropriate approach based on resources or com-
plexity of the action
Identify and estimate the incremental benefits of the action
Identify and estimate the incremental costs of the action
Compare the benefits and costs of the action
Assess the plausibility of the results
Highlight the distribution of benefits and costs and financial
impacts of the action
Integrate the assessment into other aspects of the decision-
making process. .
For example, Figure 1-2 illustrates these steps for a water quality stand-
ards action, such as a State's changing the uses designated for a river
segment. Steps in the upper portion of Figure 1-1 assemble and organize the
available data for the change in designated uses. The optional analyses of
the technical aspects of a water quality standards decision—analyses for use
attainability, site-specific criteria, and wasteload allocation—can be valuable
sources of data on the technical linkages between an action arid its effects.
By sorting the data according to whether the action's effects result in a bene-
fit or a cost to society, the State can compare, roughly, the benefits with the
costs.
The level of difficulty in the benefit-cost assessment process is dictated
by the complexity of the effects and responses to the program actions. For
example, when benefits and costs of an action are clearcut and have values
that are comparatively small, a simple qualitative assessment is in order. In
these cases, the assessment process merely describes the distribution of bene-
fits and costs—i.e., who in society receives the benefits and who bears the
costs—presents the results, and organizes them for the water quality deci-
sion. However, if a qualitative assessment reveals that potential benefits and
costs are substantial or not clearcut, a more detailed and comprehensive as-
sessment is in order, as shown by the steps in Figure 1-3. In these cases,
the practitioner must measure, value, and discount the benefits and costs and
judge the sensitivity of the results. In most instances, staff resources and
existing information can be combined for an assessment. In a few situations,
1-8
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Identify the change in
use designation for
the river segment
Obtain results of the use-
attainability assessment
for river segment
Clearcut Case
List benefits and costs and
make a determination of
the complexity of the
benefits and costs for change
in river segment
Determine as much
as feasible about
quantified and
nonquantifiable benefits
1
Determine as much
as feasible about
quantified and
nonquantifiable costs
Determine whether benefits bear a
reasonable relationship to costs
Complex Case
Determine plausibility
of results
Highlight distribution of
benefits and costs
Present results of
benefit cost assessment
Figure 1-2. Key steps in a benefit-cost assessment.
1-9
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Identify the change in
use designation for
the river segment
Obtain results of the use-
attainability assessment
for river segment
List benefits and costs and
make a determination of
the complexity of the
benefits and costs for change
in river segment
Determine nonquantifiable,
nonmonetized benefits
Complex Case
Value quantifiable
benefits using techniques
in Chapter 3
Determine nonquantifiable,
nonmonetized costs
Value quantifiable
costs using techniques
in Chapter 4
Translate benefits and cost
values into common units
using a discount rate and
appropriate time horizon
Conduct sensitivity analysis
for key variables and check
for less than fully
employed resources
Highlight distribution
of benefits and costs
Present results of
benefit-cost assessment
Figure 1-3. Key steps in a complex benefit-cost assessment.
1-10
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
outside assistance may be needed for a more detailed assessment; likely
candidates include area universities and consulting firms. Regardless of out-
side assistance, following the flow chart can ensure good quality results.
Thus, the strength of benefit-cost assessment is its ability to organize
material in a consistent manner and yet remain flexible enough to accommodate
a wide range of cases. Nonetheless, the practitioner must recognize that each
program action will introduce hew complexities requiring judgments that can be
made .based only on an understanding of the strengths and weaknesses of the
benefit-cost assessment process.
1.6 KEY CONSIDERATIONS IN AN ASSESSMENT
. Four practical problems arise in implementing general benefit-cost prin-
ciples: determining a baseline, determining the primary effects, avoiding
double counting, and using expenditures to measure benefits.
The benefits and costs of any water quality action reflect both regula-
tions already in place and specific features of the affected water bodies.
This means the baseline must be identified before the benefits and costs of a
hew action can be determined. For example, technology-based requirements,
and any variants of them, usually form the regulatory baseline for additional
water quality decisions. In cases where the technology-based requirements
have not been met, determining the baseline is difficult due to uncertainties
in the predictions of the effects that the in-place regulations will produce.
Effective enforcement is geherajly assumed for existing regulations. In an
actual assessment, practitioners will have to make judgments about these base-
line issues. If uncertainty exists in the determination of the.baseline, this
should be clearly stated and addressed later, when the plausibility of the
overall assessment is considered.
The specific attributes of a site are also important in linking benefits
with the effects of water quality decisions. For example, .swimming benefits
will not likely be significant for a river that is only a few feet deep in places
and has considerable current, no complementary facilities (such as beaches or
access points), or large amounts of barge traffic. However, swimming may be
important when adjacent parks and facilities are present and pollution is the
limiting factor, as is the case for certain river pools in the Mississippi River
in the Minneapolis area [Larson, 1981]. Similarly, the costs of achieving a
particular level of water quality will depend on site-specific water quality as
it existed before the regulatory action took effect. The analyses of use-
attainability site-specific criteria, and wasteload allocation, any one of which
may be performed as an optional part of water quality standards decisions,
can be a valuable source of technical information.
Another important distinction is between primary and secondary benefits
and costs. Primary benefits and costs arise directly from the action, while
secondary benefits and costs follow the impact of the primary ones. Only
primary benefits and costs should be included in an assessment, because link-
ages are often too imprecise to make even a rough determination of secondary
benefits and costs. For example, .while increased recreation activities and
7-11
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
enhanced ecological diversity are among the primary benefits of a water quality
improvement, the increased revenues to providers of recreation equipment,
for example, are simply expenditures—secondary benefits-- not primary bene-
fits. When these expenditures are included, the opportunities for double
.counting increase substantially. If increased receipts of recreation equipment
suppliers are added to willingness to pay, then that part of users' willingness
to pay is double counted. In effect, including secondary benefits in an as-
sessment opens up the assessment to the same suspicions that plague some
applications of traditional cost-benefit analysis.
The distinction, between primary and secondary benefits is important in
the solution of another problem that arises in a benefit-cost assessment: the
use of an expenditure approach to measure benefits. The expenditure ap-
proach adds up an area's recreation-related expenditures. This reflects the
costs of recreation, not an individual's willingness to pay for recreation. For
example, the approach would include the costs of the fishing gear itself—
amounts that are costs and not benefits; In addition, it does not count the
difference between the maximum an individual would pay and the amount he
actually pays—in technical terms the consumer surplus. In effect, the ex-
penditure approach includes some costs on the benefits side of the ledger and
excludes other benefits entirely.
Total recreation expenditures may be useful in identifying some of the
effects on a community's economic activity (e.g., increased sales tax receipts
or recreation-related employment). Even in this limited use, however, the
use of total expenditures omits important flows of funds out of the community
to pay for goods externally produced. Confusion on these points often re-
sults because it seems logical that expenditures should be benefits. However,
expenditures are costs and benefits; they are not al| benefits. Both double
counting and miscounting occur when this approach is used.
1.7 SUMMARY
Benefit-cost assessment:
Applies a formal dose of common sense to evaluating water
quality regulations and programs.
Provides a flexible approach for organizing the information
needed to make water quality decisions.
Enhances but does not supplant the value judgments of
decisionmakers.
Uses.society as the basis for accounting benefits and costs.
Focuses on individual willingness to pay and opportunity cost
to measure benefits and costs, respectively.
Concentrates on primary benefits and costs.
1-12
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
1.8 GUIDE TO HANDBOOK
This handbook is organized into six chapters. Chapter 1 is the intro-
duction to benefit-cost assessment. Chapter 2 shows how to include intangi-
bles arid distribution effects in an assessment and considers the question of
discounting benefits and costs. Chapter 3 describes specific methods for esti-
mating the benefits of water quality programs, including techniques for devel-
oping the monetary components of benefits needed in complex cases. Chap-
ter 4 gives the basics of estimating costs, focusing on the incremental costs
of water quality regulations. Chapter 5 describes a sensitivity analysis as a
guide to a plausible assessment and highlights methods of presenting the re-
sults of an assessment. Chapter 6 illustrates benefit-cost assessment prac-
tices with simple, moderately difficult, and complex causes to reflect different
types of water quality decisions.
1-13
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
CHAPTER 2
ISSUES IN A BENEFIT-COST ASSESSMENT
2.1 INTRODUCTION
How can intangible benefits and costs be included in a benefit-cost
assessment? What does it mean to discount benefits and costs? Are there
rules of thumb for discounting? Does benefit-cost assessment ignore the
distribution, of benefits and costs?
In practice, the positive and negative effects of a program action occur
at different points in time, affecting households and firms over a number of
years. In many assessments the need or ability to monetize benefits and
costs may be small, or limited resources may preclude monetizing. Clearly, a
need exists for a logical approach to intangibles and for a convenient way to
include them in an assessment.
For those assessments where benefits and costs are monetized, two main
"adding up" issues arise: discounting and distribution. Discounting pro-
vides a consistent basis for adding benefits and costs over time. |t is one of
the most complex and controversial issues in an assessment. Similarly, simply
adding benefits and costs over people or firms may hide important issues.
To explain how a benefit-cost assessment addresses these important
Issues, Section 2.2 discusses intangible benefits and costs and uses arrays,
or tabular, displays, to feature them in the assessment. Section 2.3 briefly
describes how to measure impacts on firms and households, and Section 2.4
illustrates how the distribution of benefits and costs can be included in an
assessment. Section 2.5 describes the discount rate, its role in a benefit-
cost assessment and key issues in selecting a discount rate. Finally, Sec-
tion 2.6 summarizes the chapter's major points.
2.2 INCLUDING INTANGIBLES IN A BENEFIT-COST ASSESSMENT
Introduction
The uses prescribed in the Clean Water Act for the water quality stand-
ards are likely to provide intangible benefits relating to enhanced species
diversity and ecological habitats and improved aesthetics. By its incommen-
surability, this type of benefit presents problems for determining the net
benefits of a use designated under the standards program--or for any water
quality program. The types of benefits or costs that comprise the intangibles
group change over time with improvements in valuation techniques. For
example, the travel cost technique estimates willingness to pay for recreation
benefits that initially were treated as intangibles. This section provides a
2-1
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
method for including intangibles in a benefit-cost assessment. Volume II will
contain more details on intangible benefits.
A recommended method uses a system of tabular displays, or arrays, to
present both tangible and intangible benefits. These arrays are tailored to
fit the nature of the assessment being conducted. The first array simply
lists and describes the benefits and costs. A second array presents mone-
tized values only for those benefits and costs for which monetization is almost
always accepted, with the remaining benefits and costs being listed, de-
scribed, and quantified to the extent possible. These values are based on
individuals' willingness to pay and opportunity cost as measured by the tech-
niques presented in Chapters 3 and 4. The final array presents as many
benefits and costs in monetized terms as possible, with the other benefits and
costs being listed, described, and quantified.
Example
Suppose a State is considering adding the fish and wildlife use to a
stream that is currently designated for agricultural and industrial uses. A
system of arrays for a benefit-cost assessment of this change in use designa-
tion is illustrated in Tables 2-1, 2-2, and 2-3.
The first array (see Table 2-1) lists and describes all the benefits and
costs in qualitative terms. In cases where issues are clearcut, this array
would, by itself, provide information sufficient for making the decision. As
issues become more complex, additional arrays are essential for information
sufficient to make the decision.
The second array for this example (see Table 2-2) presents monetary
values for benefits and costs for only those categories that almost all practi-
tioners agree can be monetized. This supplemental information in the second
array clarifies the issues in the assessment by focusing attention on the
nonmonetized values. This array shows that the low end on the range of
monetized recreation benefits is exceeded by the high end on the range of the
costs and that the action produces nonmonetized benefits. The decisionmaker
would have to determine how the nonmonetized aesthetic benefits and ecologi-
cal diversity influence the net result given the overlap in the range estimated
for benefits and costs.
i
The system of arrays shows how a benefit-cost assessment can reduce
the dimensions of a complex issue to focus the decisionmaker's attention on
the most difficult aspects. The last array (see Table 2-3) shows the mone-
tization of as many benefits and costs as possible. The range of monetized
benefits is estimated to be $17 million to $37 million, with additional nonmone-
tized benefits attributable to the enhancement of the ecological diversity.
The estimated costs of attaining the additional use designated range between
$9 million and $14 million.
Several features of this example call for additional discussion. The
estimated monetary values for aesthetic benefits move the benefit range to a
level at which benefits exceed both the minimum and maximum estimates of the
2-2
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Table 2-1. Benefits and Costs of Attaining Fish
and Wildlife Use: Array 1
Description of Benefits
1. Provide an additional resource that can support swimming, fishing, and
recreation near water in a metropolitan area with only limited substitutes
available.
2. Improve the aesthetic value for users of the resources services, such
as recreators or property owners near the stream.
3. Improve the aesthetic vatue for residents of the area based on possible
use in future or just from knowing the stream is cleaner.
4. Enhance the ecological diversity of the stream area by providing an
improved habitat for fish species and wildlife from surrounding areas.
However, none of these species is unique or endangered.
Description of Costs
1. The residents of the city will require advanced treatment for their
wastes.
2. Three industrial dischargers will have to modify their waste treatment
operations.
Table 2-2. Benefits and Costs of Attaining Fish
and Wildlife Use: Array 2
Quantity
Monetary value
(million $,
present values)
Types of Benefits
1. Fishing, swimming, recreation
near water
2. Improved aesthetics for
users--recreators and prop-
erty owners near stream
3. Improved aesthetics for
nonusers--vatue to residents
in area from knowing stream
is clean should they use it
or from just knowing it is
clean
4. Enhanced ecological diversity
1 million visits
10 to 30
Types of Costs
1. Advanced treatment for
municipal wastes
2. Advanced treatment for
industrial dischargers
10 new fish species,
smallmouth bass and
others; 1,000 acres
of improved wildlife
habitat; no unique
species are provided
1 new plant
3 additional treat-
ment operations
8 to 10
1 to 4
2-3
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Table 2-3. Benefits and Costs of Attaining Fish
and Wildlife Use: Array 3
Quantity
Monetary value
(million $,
present values)
Types of Benefits
1. Fishing, swimming, recreation
near water
2. Improved aesthetics for
users--recreators and prop-
erty owners near stream
3. Improved aesthetics for
nonusers
4. Enhanced ecological diversity
1 million visits
10 to 30
5
Types of Costs
1. Advanced treatment for
municipal wastes
2. Advanced treatment cost
for industrial dischargers
10 new fish species, Not monetized
although no unique
species are provided;
1,000 acres of improved
wildlife habitat.
1 new plant 8 to 10
3 additional treat- 1 to 4
ment operations
costs. By monetizing these benefits with a contingent valuation survey
(discussed in Chapter 3), the practitioner can show that the minimum benefits
exceed even the highest cost. This makes a strong case for adding the use.
In addition, the array in Table 2-3 shows nonmonetized benefits that would
increase the total value of the benefits even more. This example illustrates
the case for a river segment with large recreation potential that justifies the
extra cost involved in carrying out the detailed assessment. River segments
that have intermittent flows or whose entire flow is effluent would have low
recreation potential and would not require such a detailed assessment.
To show how the system of arrays can present assessment results, the
discussion of this example concentrates on the efficiency aspects of the use
designation. As noted in Section 2.4, however, the decision process .also
should consider information on distribution effects.
2-4
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
2.3 COST IMPACT MEASURES
This section describes the general approach and some specific measures
for assessing the cost impacts on communities and industries of meeting water
quality regulations. This handbook does not advocate the use of any one
measure; rather, it emphasizes that impacts be evaluated. Overall, the
objective is to determine the incremental effect of compliance costs on earn-
ings, production, and employment in the affected locality. However, the
financial ability of a community or industry to absorb these costs is also
important.
For each major impact category, the following sections describe one or
more measures with varying degrees of sophistication, data requirements,
estimation methods, feasibility, and accuracy. By no means :are these meth-
ods the.only way to proceed.
Assessing Household Impacts
The share of publicly owned treatment works (POTW) costs allocated to
households in the form of higher sewer or water charges is assumed to be
borne directly by those households, reducing annual income by the amount of
the total annual costs. This implies that households cannot pass oh these
costs by increasing their wages. Together with data on household income,
total costs of compliance and community indebtedness are used to develop
measures of the ability of households to bear these costs—i.e., how these
costs affect income arid indebtedness. The discussion of impact measures is
kept brief here because an EPA document, the.Financial Capability Gu|debook,*
provides a detailed guide to community financial capability analysis.
Household impact measures are divided into two major types: ability to
pay and ability to finance. Ability-to-pay measures focus on the ability of
the residents to bear the costs of water quality improvements, regardless of
the current financial status of the community. Ability-to-finance measures
focus on the ability of the community to finance the costs of compliance. The
Financial Capability Guidebook develops 11 key indicators used to judge the
ability to bear the impacts specified by these measures.
Ability to Pay
The measure of ability to pay is the ratio of compliance costs to median
household income. Compliance costs are defined as total annual costs: the
sum of annualized capital costs, operation and maintenance costs, and contri-
butions to contingency funds. Focusing on values for the median household
is convenient, but particular situations may call for a more detailed examina-
tion of the distribution of income or wealth in the community.
*Municipal Finance Officers Association and Peat, Marwick, Mitchell &
Co., Financial Capability Guidebook (Draft), prepared for EPA, Office of
Water Program Operations, Washington, D.C., May 1982. Contact the OWpO
at EPA for further information about this document.
2-5
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
The Income measure may be used to estimate two kinds of cost impacts:
the incremental costs of a particular water quality program and the total costs
of compliance associated with all treatment. In some cases, total costs may be
useful because the practitioner is interested in a view of cost impacts overall.
For specific water programs, however, a focus on the incremental costs of
decisions .that move water quality beyond that resulting from the statutory,
technology-based controls can reveal the cost impacts due to the individual
program.
Financial Capability
The financial capability of a community has a significant effect on its
ability to raise additional funds in the bond market. If a community has high
indebtedness or low tax revenues, it will have a. lower bond rating and face
higher costs of capital. The impact measure is the total outstanding debt of
the community before and after the project being considered.
Assessing Industrial Impacts
A firm's compliance cost may arise either from direct costs of treatment
facilities or process changes or from its share of POTW costs. In turn, these
costs lead to changes in profitability, output, and employment and may result
in partial or complete closures. Of course, an accurate measure of impact
requires some estimate of a business cash flow, but this estimate may be
difficult to obtain unless the companies concerned cooperate. Thus, a large
tradeoff exists among the measures .with respect to feasibility and accuracy.
Table 2-4 summarizes the proposed impact measures, data sources, and their
availability and reliability. The measures will be described in greater detail
in Volume II.
Assessing Changes in Employment, Output, and Prices
Changes in output and employment in response to treatment costs are
important because they give rise to indirect impacts. If workers are unem-
ployed, they reduce their spending; if a plant reduces output, its demand for
inputs from supplying firms slackens, with indirect repercussions on commun-
ity income and employment. Even if estimating indirect impacts is infeasible,
the direct effects of compliance costs on community employment and income are
useful for assessing the equity implications of a regulatory action--"who is
affected?1'
As a rule of thumb for assessing the effect of regulatory actions, firms
usually cannot pass through treatment costs by raising prices. This rule
normally holds for water quality standard actions because they are site-
specific and may affect only certain businesses. Although the rule may not
hold for some regulatory actions, it is difficult to predict under what circum-
stances firms might be able to pass through costs.
2-6
-------
1
1
1
1
1
1
1
1
1
IB
Table 2-4. Summary of Industrial Impact Measures
Measure
Profitability
(including closure)
1. Cost/sales ratio
2. Cost/production
cost ratio
3. Net cash flow
4. Rate of return
5. Net present value
6. Company solvency
Reductions in employment
and output
1 . Due to closure
2. Due to output
reduction
Price changes expected
to be small
Source
Production and
price estimates
or public data
bases
EPA economic
impact analysis
Plant financial
data
Plant and company
financial data
Plant and company
financial data
Company financial
data •
Plant data
Plant data, engi-
neering report
Availa- Relia-
bility bility
High Low
/
High Low
Medium Medium
Low High
Low High
Depends Depends
on size on size
Medium Medium
Medium Medium
2.4 WHAT TO DO ABOUT DISTRIBUTION: PROBLEMS IN
ADDING UP OVER PEOPLE
I ntroduction
A net benefits estimate does not evaluate projects based oh the distri-
bution of net benefits. Rather,
the evaluation is based on efficiency criteria,
U.S.
EPA Headquarters Library
Mail code 3201
2-7 120° Pennsy/vanfa Avenue NW
Washington DC 20460
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
which show how to allocate society's resources to maximize well-being.*
Although the weights assigned to individual recipients of the benefits and
costs of a project are treated equally, the distribution of net benefits can be
described for the 'policy under evaluation. In these descriptions, benefits
and costs are separated according to the economic agents affected. For
example, one might classify households by income group, or firms by indus-
try, and evaluate each group's share of the net benefits. Then, the overall
benefit-cost assessment can account for distribution.
Some attempts have been made to explicitly include equity (in terms of
the effects of the project on different income groups) in benefit-cost assess-
ment. They are not uniformly accepted. All of the weighting schemes are
based on the premise that the extra utility or satisfaction derived from an
extra dollar of income declines as income increases. Thus, redistribution of
income will lead to increases in total utility or satisfaction for society.
There are flaws inherent in any weighting scheme for benefit-cost as-
sessment. The most difficult one to overcome is that it is hard to get society
to agree on the appropriate weights. Equity in income is only one possibil-
ity; regional and racial equity are others. In the final analysis, the weights
are simply attempts to "guesstimate" the decisionmaker's preferences. This
does not imply that distribution information should not be developed. Rather,
it suggests that the tradeoff between efficiency (as measured by the aggre-
gate net benefits) and various types of equity considerations (as reflected in
the distributions of these net benefits among economic agents .under different
classifications) is unlikely to be capable of being assigned a fixed relation-
ship. Ultimately, the importance of distributional issues will depend on the
decisionmaker's judgment,
' Two examples can be used to illustrate how distributional information has
supplemented the conventional net benefit information in a benefit-cost as-
sessment.
Example I .
Suppose an:improvement in water quality will provide. $10 million a year in
net benefits. The distribution issue is to determine who will receive these
benefits. The most commonly used method is to array the benefits by the
shares that will accrue to different income groups, as shown below:
Income ($)
Less than 10,000
10,000 to 20,000
20,000 to 35,000
More than 35,000
Percent of net benefits
20
30
35
15
*Efficiency criteria indicate both the cheapest way of achieving a partic-
ular level of water quality and what level of water quality makes sense given
competing uses of resources.
-------
r
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
This change would favor lower income groups: half of the net benefits
accrue to people with incomes of $20,000 or less, and over three-fourths
accrue to people with incomes of less than $35,000 a year. By quantifying
and monetizing benefits and costs, benefit-cost assessment provides a clear
picture of the distribution effects of the change. The distribution of any
intangible benefits and costs also should be considered in the decision.
Example II
Consider the same situation as in the previous example with $10 million
in net benefits from the change in water quality. Table 2-5 arrays the
distribution of benefits and of project costs. Rather than arraying by income
groups, the categories in Table 2-5 break down the distribution of benefits
and costs over broad groups in society to illustrate another way that distri-
bution effects can be highlighted.
Table 2-5. Distribution of Benefits and Costs
Benefits: Who Receives?
Users of river for recreation
People who receive enhanced aesthetics values for river
Downstream users for municipal water supplies
Downstream companies who use water for industrial processes
Costs: Who Bears?
Residents who incur higher sewer and water bills because of advanced
treatment requirements for wastes
Stockholders of companies who have to install new equipment or change
production processes to meet the standards
Consumers who purchase products whose prices are increased as a re-
sult of companies' compliance
Summary: Distribution
Information on distribution effects of water quality programs is an essen-
tial ingredient in a benefit-cost assessment. It can be described with either
summary measures like income group shares, or simply listed in narrative
form.
2-9
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
2.5 DISCOUNTING FUTURE BENEFITS AND COSTS: ADDING UP
OVER TIME
One of the most crucial issues in a benefit-cost assessment that relies on
monetized benefits and costs is the selection of the appropriate discount
rate—an interest rate used to translate dollar amounts of benefits and costs
occurring in different years into a common unit of comparison, usually a pres-
ent value. The discount rate is a positive number because individuals prefer
immediate consumption and the associated immediate satisfaction to future con-
sumption and the corresponding future satisfaction. Then, to persuade indi-
viduals to give up immediate consumption in exchange for future consumption,
the level of consumption must be increased at that later date. This increase
is an opportunity cost and is .demonstrated, for example, when companies pay
interest, or share future profits, to take advantage of current investment op-
portunities. Preference for satisfaction now rather than later—technically
known as the positive rate of time preference—is demonstrated when someone
installs new carpet on a time payment plan or finances a new car. The satis-
faction or utility from the carpet or car is gained now at the expense of a
financing charge.
There are five key concepts in determining discount rates:
Social rate of time preference: the rate at which society is
willing to exchange present consumption for future consump-
tion.
Consumption rate of interest: the rate at which individuals
are willing to exchange present consumption for future con-
sumption.
Marginal rate of return on private investment: the incremental
return on the last unit of investment by a private firm.
Opportunity cost of public investment: the cost of a govern-
ment investment measured in terms of forgone private consump-
tion or investment.
Risk: the degree to which investment in a public project will
affect the variation in the outcome of all public investment.
While a large share of the costs of meeting a water quality standard
occurs in years immediately after a standard is .set (e.g., firms invest in new
treatment processes, and cities construct advance waste treatment plants),
benefits will not accrue until after the new plants and processes are in place.
These benefits may accrue for 50 or 100 years—a period over which dollar
amounts of both benefits and costs will vary greatly. The pattern of dis-
counted net benefits often will look very different if a high discount rate is
used rather than a low one. As this example shows, assessing benefits and
costs of any water quality program requires an appreciation of the basic
principles underlying the definition and the selection of an appropriate dis-
count rate, an understanding of the empirical implications of discounting, and
a practical knowledge of how to work with discounting techniques.
2-10
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Time Preference: What Is It?*
The use of discount rates in benefit-cost assessment can be explained by
viewing discounting issues within the context of an ideal market economy, the
characteristics of which include perfect competition in all markets, complete
certainty in decisions., no transaction costs, no taxes, and no limitations on
any credit market. In such an economy, all goods are priced at the oppor-
tunity costs of the inputs used to produce them, and individuals and busi-
nesses are able to borrow or lend, subject to their ability to repay, as much
as they desire at the market rate of interest, which is determined by the
demand and supply of loanable funds. .
In hopes of obtaining future earnings, businesses in this ideal world
would invest funds to. the point where their extra benefit equals their oppor^
tunity cost. In this economy, the market interest rate will be the opportun-
ity cost of capital for the firm. Consequently, optimizing behavior by each
business and efficiently working markets will ensure that the market interest
rate will be equal to the marginal return on investment. The firm could not
rearrange its investments and improve its long-term profit picture. At the
same time, individuals would arrange their consumption and savings such that
their own marginal rate of time preference (also known as the consumption
rate of interest) would equal the market rate of interest. In this ideal case,
the final outcome is that market forces create an equilibrium in which an
individual's consumption rate of interest and a firm's marginal return on
investment are the same because both correspond to the market rate of inter-
est. This equilibrium ensures an efficient allocation of resources over time.
If these equalities were not maintained (for example, if the consumption rate
of interest were less than the marginal return on investment), an individual
could improve his welfare (by consuming less, saving, and earning a return
that permitted greater consumption in the future).
Introducing public investments—such as those mandated by water quality
programs—requires that the resources supporting them displace either private
consumption or private investment. An efficient allocation of resources means
that these investments earn a return at least equal to the marginal return on
capital or the consumption rate of interest that would be required for these
alternative uses (i.e., private investments or consumption). If it is also
assumed that all individuals are alike with respect to factors determining their
rates of time preference, society's overall rate of time preference should equal
the market rate.
The implication for selecting a discount rate for the water quality pro-
grams in an ideal society is that either the social rate of time preference or
the opportunity cost of capital would be appropriate because they are the
same—i.e., both are equal to the market rate of interest. Unfortunately,
when the assumptions of the ideal case are relaxed, the two rates diverge.
These divergences explain why the selection of a discount rate for public
sector investments has been such a difficult and often controversial issue.
This section draws extensively oh Lind et al. [1982]
2-11
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
How To Determine the Djscount Rate in a Less Than Ideal World
The selection of a discount rate for a benefit-cost assessment of a water
quality program must be accomplished in a world considerably different from
that of the ideal economy. Some of the most important divergences from the
ideal economy can be attributed to the following factors:
The tax on corporate income, which drives a wedge between
the private rate of return and the rate of time preference. A
higher rate of return on private investment is required to off-
set the effects of income taxes, which cause the divergence
between social and private rates.
The dependence of future generations on decisions made by
present generations. This dependence gives rise to a "public
good"--the welfare of the future generations—that may not be
included in the decisions of the private market.
Private markets, which may be out of long-run equilibrium
with an immediate discount rate different from the appropriate
long-term rate.
The determination of the appropriate private market rate,
which is difficult because there are numerous capital markets,
each with its own interest rate.
» Public investment dollars, which do not necessarily displace
private investment dollars but may instead use tax revenues
that displace current consumption in the private markets.
Attempts to reconcile these divergences have created a complex literature
on the criteria for selecting an appropriate discount rate for public invest-
ments or regulatory evaluations under different circumstances. Lind [1982]
has distilled this literature, concluding that the rationale for discount rate
selection should be based on the full opportunity costs of capital.
Lind's approach does not ignore the potential divergence between socie-
ty's social rate of time preference and the market rate of interest. Rather,
he suggests that the social rate of discount be set equal to the social rate of
time preference and that the shadow price of capital be used to adjust for the
full opportunity costs of capital. The shadow price of capital is defined as
the present value of the future stream of consumption benefits associated with
$1 of private investment discounted at the social rate of time preference.
A benefit-cost assessment to evaluate government investment decisions
based on this approach considers the implications that investments have for
consumption over time. The basic question to be answered is, "What does
public investment displace?" To the extent public investment displaces pri-
vate investment, that portion of the costs of the public project should be
valued at the shadow price of capital. That is, the costs of this portion of
the investment are valued in terms of the consumption forgone. When the
2-12
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
forgone consumption due to displaced investment is added to the balance of
the costs (i.e., those displacing immediate consumption), all costs have been
converted to their equivalent losses in private consumption. Similarly, bene-
fits that lead to increased private investment should also be adjusted to
reflect their potential to yield future consumption streams. The adjustment is
to multiply that share of the private investment by the shadow price of capi-
tal. This procedure adjusts benefits and costs at each point in time and
expresses them in terms of the equivalent amount of .consumption that could
be obtained.
Table 2-6 .shows estimates of various parameters so that a practitioner
could use this general procedure in a benefit-cost assessment. The estimated
value for the marginal return on private investment in Table 2-6 is 10 per-
cent. This approximation is based on judgment and the empirical relationship
that the estimated average return, adjusted for inflation, is between 10 and
15 percent. More accurate measurements of the costs of capital are hindered
by the Inconsistency between accounting data and economic concepts, and by
implicit adjustments both for risk in the returns to capital data and for firms'
inconsistencies in following established procedures to make capital budgeting
decisions.
Based on depreciable assets data for 1973, 1974, and 1975, the length of
the typical private investment is estimated at 15 years, with a range of 10 to
20 years. Based on empirical work on consumption and savings, the marginal
propensity to save is assumed to be 0.2.
Table 2-6. Summary of Final Description Factors Influencing
Shadow Price of Capital
Shadow
price of
capital
1.62 to
2.57a
1.9b
Marginal
propensity
to save
0.2
0.2
Marginal
return .on
private capital
(%)
10
10 to 15
Social rate
of time
preference
(%)
2
2 to 6
Length of typical
private investment
(years)
10 to 20
15
This row is based on Lind's Table 4 comparing the shadow price of capital
under a range of assumptions for the social rate of time preference, marginal
return on private investment, and length of typical private investment.
This row is based oh Lind's discussion on pp. 101-102 of the unpublished
manuscript. The returns on private capital are pre-tax returns. The
shadow price of 1.9 is the central value associated with the variations in
each of the parameters involved. The range of values was 1.65 to 2.15.
The range of social rates of time preference are reported to indicate that
they would be consistent with this shadow price.
2-13
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
In the procedure, the social rate of time preference is set equal to the
consumption rate of interest for individuals. Unfortunately, there is no
unique rate of interest in the real economy. But it is possible to gain some
insight into individuals' rates of time preference from their savings and
investment decisions. For example, the real after-tax average rate of return
on treasury bills (a safe investment available to many people) over the period
1926 to 1978 was -0.5 percent for an individual in the 20-percent tax bracket,
and the return on a mutual fund containing "average market" stocks totaled
4,6 percent. For individuals, the real rate of return must lie somewhere
between the riskless treasury bill rate and the stock market equity returns. ,
Adjusting for the effects of unanticipated inflation shows that the mar-
ginal rate of time preference must be in the range of -2 and 5 percent, with
the average close to 0. Regardless of the actual point estimate selected, it
will be considerably different from the 10 percent real rate of discount recom-
mended by the Office of Management and Budget (OMB) for public, investment
projects.*
The shadow price of capital can be calculated from the estimates of the
other parameters following Lind's procedures. Specifically, Lind estimates
that the most likely value is within the range between 1.65 and 2.15. This
shadow price is then used as described above to convert the benefits and
costs into consumption equivalents.
Including Risk in an Assessment of Discount Rates
So far the implications of uncertainty for determining the discount rate
have been implicitly ignored. Since the levels of benefits and costs are
uncertain, there should be adjustments for the probabilities that a particular
level of each will occur. Depending on its source and nature, the uncer-
tainty can either be addressed in this choice of a discount rate or be directly
reflected in the measurement of benefits and costs. Alternatively, some com-
bination can be attempted. Each of these alternatives will be considered in
Volume II.
The main conclusion that can be drawn from the risk studies covered in
Volume II is that the characteristics of the public investment project, together
with the relation between the variability in the public investment and the
variability in national income, are crucial factors in determining whether a
riskless or risk-adjusted discount rate should be used in an assessment of
public investments. The important relationship that must be explored is
whether water quality program investments increase the variability in national
income. For most of these applications, the effects of risk will be small, and
a riskless rate can be used.
*lt is difficult to interpret this rate as an estimate of the social rate of
time preference. It may well be the equivalent to a recommendation that all
projects be discounted at the marginal rate of return on private investment.
Additional confusion is added by the Water Resources Council [1979] guidelines
which tie the discount rate to an index. The current rate in this procedure
is 7-7/8 percent.
2-14
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
What Are the Empirical Implications of the Discount Rate .Issues?*
One of the reasons for the controversy concerning the use of discount
rates is that the empirical implications of the discount rate can have a sub-
stantial influence on the outcome of the benefit-cost assessment. Fox and
Herfindahl [1964] reevaluated Federal water projects, previously evaluated at
a 2-5/8 percent discount rate, at new discount rates of 4, 6, and 8 percent.
Nine percent of the projects that previously had benefits exceeding costs at
2-5/8 percent experienced the opposite result with a 4-percent discount rate;
64 percent experienced the opposite result at a 6-percent rate; and 80 per-
cent experienced the opposite result at an 8-percent rate. The implications
are quite clear: Most of the projects .had costs exceeding benefits at the
higher discount rate, and all had positive benefits at the low 2r5/8 percent
rate.
Many water program regulations or projects will have benefits that will
accrue 10, 20, 30, or even 100 years in the future and costs that could be
substantial during initial periods. For example, suppose a water quality
project requires a $1 million investment in 1982 and will provide $50 million in
recreation .benefits at the end of 50 years. The net present values at differ-
ent discount rates would be as follows:
Discount rate
5 percent
8 percent
12 percent
Net present value
3,336^186
6,606
-826,990
As shown, the discount rate is crucial in determining the ultimate assessment
of benefits and costs. This emphasizes the importance of the sensitivity
analysis recommended in Chapter 5, which shows that in some cases there will
be positive net benefits regardless of the discount rate, while in others the
outcome of the assessment is very sensitive to the discount rate applied.
Another important empirical distinction in discounting is the difference
between real and nominal rates of discount. The difference is the expected
rate of inflation. Most benefit-cost analyses are conducted using constant
dollar values for the benefits and costs. In these cases 'the real rate of
discount should be used.
Benefit-cost assessments have employed real discount rates ranging from
0 to 4 percent, while the nominal rates have ranged from 8 to 16 percent.
The higher end of the scale for nominal rates represents the influence of
recent high levels of inflation and market interest rates. The .difference
between the real and nominal rates is quite substantial and indicates why it is
important not to mix the two in a benefit-cost assessment. For example, com-
pare the implications of a real discount rate of 2 percent and a nominal rate
of 10 percent (inflation is expected to be approximately 8 percent). With the
*This discussion is adapted from Just, Hueth, and Schmitz [1982].
3-15
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
nominal rate used as the discount rate, society would be indifferent between
$1 now and $13,780 in 100 years. However, if a real rate is employed, the
difference would be $1 now versus $7.24 in 100 years. Thus, although it
seems low, given the high interest the economy is currently experiencing, a
real discount rate of 2 to 4 percent may make considerable sense in a benefit-
cost assessment where the long-term perspective is essential.
The Simple Mechanics of Discounting
This section offers a brief review of the mechanics of discounting. The
reader is urged to consult a finance text for present value tables and more
detailed discussions on discounting.
The discounting problem in a benefit-cost assessment is how to translate
benefits and costs occurring .in different periods into a common basis for
comparison. The most frequently used basis is present value, which is
defined as the amount of money at the present time that some future amount
is worth. Discounting is the process of computing the present value of a
future stream of dollars.*
Consider a simple example that might arise in the assessment of a water
quality standards program. Suppose a State is considering changing the
presently unattained use designation for a river segment from "fish and wild-
life propagation" to "limited warmwater fishery." In this case, there will be a
loss in potential benefits, as well as cost savings from the forgone pollution
control investment for cities or industries. The monetary values for the
benefits and costs associated with this decision are as follows:
Year Benefits forgone Cost savings
1982 40,000 $100,000
1983 40,000 10,000
1984 40,000 10,000
The discounting problem is:
1. Select the appropriate discount rate.
2. Translate future benefits and costs into present values for
comparison.
The discounting formula for this procedure is:
1
P.V. =
(HO*
*For simplicity, assume all dollars accrue at the end of each year so
there is no need to account for differences within a year.
2-16
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
where
P.V. = the present value factor for either benefits or costs
i = the discount rate
t = the time period.
The discount factor P.V. is multiplied by the benefits or costs foK each
period in the planning horizon; then the results are summed. Suppose that
i = 4 percent with the monetary values for benefits and costs from above.
The discounting calculations are the following:
(1982)
(t=0)
an nnn ±
(1983)
(t=1)
(1984)
(t=2)
Benefits forgone = 40,000 +
Cost savings = 100,000 +
Benefits forgone = 40,000 + 38,462 + 36,982
Cost savings— 100,000 + 9,615 + 9,246
Net cost savings = $3,317 .
$115,444
$118,861
The change of the use designation will yield a small positive cost savings. In
this example, the forgone benefits end in 1984. In most cases, they would
continue into the future for whatever time horizon is selected for the assess-
ment.
i
If the benefit (or cost) stream is constant each year at A for the full
life of the project (assumed to be n years), the formula for the present value
can be rewritten as follows, often referred to as the present value of an
annuity:
1
P = A
1." (1+0
n
where
P = present value
A = annual amount
i = discount rate.
Present value may also be determined by using the tables in a finance or
accounting text. Another .variation on the discounting mechanics is to trans-
2-17
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
late amounts into annual values that can be compared. The formula for this
calculation, often referred to as the uniform-series capital recovery factor,
is:
i(l-H)
A = Pj
Discounting: A Summary Review
While the arguments summarized above require the use of judgment for
each new situation, general guidelines do exist. For a wide range of water
quality programs, the social rate of time preference would range from 2 per-
cent to a maximum of 6 percent. The recommended procedure is to consider
the implications of each of these values for the discounted net benefits of the
decision.* If the present value of net benefits is positive and the project
decision remains unchanged, it is unnecessary to further consider the techni-
cal issues affecting the selection of one value in this range. However, for
those cases where the value of net benefits (i.e., positive versus negative) is
affected by the discount rate, a more refined selection is recommended that
considers the practical implications of Lind's analysis. This process requires
answering four key questions:
1. What are the sources of the public investment resources? Are
they tax revenues that can, in principle,, displace private
consumption or investment?! In addition, what are the likely
portions coming from each source? Answers to these questions
will affect the importance of estimating the shadow price of
capital.
2. How large are the private investments required by the water
quality action? If they are large, adjustment by the shadow
price is likely to be warranted to reflect the full opportunity
costs of these investments.
3. What is the nature of the risk associated with the investment
and its relationship to overall economic activities? The answer
to this question provides a basis for judging whether an
adjument for risk should be made in evaluating the project. If
most water quality investments increase the variability in eco-
nomic activity for a State (or the county as a whole), then the
selection will tend to be at the higher end of the range for the
social rate of time preference.
*ln some cases, legal restrictions mandate the use of a specific rate;
e.g., advanced treatment applications require the applicants to use the Water
Resources Council's rate of 7-7/8 percent.
tUser fees should be regarded as payments for services provided and
therefore do not displace private investment.
2-18
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
4. Are the sources of finance for the project known in advance
(i.e., Federal sharing of a local project's cost)? Since one
objective of benefit-cost assessment is to improve the overall
allocation of resources, any Federal share of the costs should
be treated in the same way as the local share, with consid-
eration given to the full opportunity costs of the funds used.
It is impossible to recommend a single rate of discount as relevant for all
situations. Each decision may well have special attributes that will need to be
reflected in the selection. It is important not to let the technical considera-
tions involved in defining the appropriate discount rate become overwhelming.
For most purposes, the 2 to 6 percent range of values will be all that is
necessary.
2.6 Summary
Intangibles .should be viewed from society's willingness-to-pay
benchmark even though they are incommensurable.
Arrays or tabular displays are useful exposition tools for
intangibles.
Household impacts can be measured according to ability to pay
or ability to finance.
One measure of ability to pay is the ratio of compliance costs
to median household income.
Ability to finance is reflected in a city's bond market rating in
the financial community.
Industrial impacts can affect profits, output, and employment.
Distribution can be highlighted in arrays showing who receives
and who bears.
Discount rates should be selected carefully. Shadow price of
capital should be considered in gauging the full opportunity
cost of public investment.
A sensitivity analysis should be performed for the effects of
selecting the. discount rate. Several discount rates should be
tried.
The implications of risk should be considered for the discount
rate. The relationship between the variability of investment or
outcomes of regulatory policy relative to variability in national
income should be considered.
2-19
-------
I
Net benefits should be determined on a present value basis.
Formulae and tables should be used.
• * Real and nominal discount rates should not be mixed.
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
2-20
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
CHAPTER 3
MEASURING THE BENEFITS OF WATER QUALITY PROGRAMS
3.1 INTRODUCTION
What are the benefits of water quality programs? What are the basic
benefit principles? What methods are available to measure benefits? What are
the key assumptions and features of these methods? What case studies are
available to illustrate the use of these1 methods?
While benefit-cost assessments that use only qualitative information may
suffice for clearcut water quality decisions, quantitative information can simpli-
fy more difficult decisions, by reducing the complexity of the issues and clari-
fying the central issues. This chapter provides a framework for gathering
and organizing this information to measure the benefits of water quality pro-
grams.
This chapter briefly reviews the concept of benefits and practical ap-
proaches for measuring them. In particular, it discusses the conceptual
issues involved in estimating the benefits of water quality improvements, de-
scribes techniques for measuring different types of benefits, and presents
case studies that show how different practitioners have employed them. Spe-
cifically, Section 3.2 summarizes benefit categories, Section 3.3 highlights
general issues in selecting a benefits estimation methodology, and Section 3.4
describes approaches for measuring household benefits. Section 3.5 discusses
business benefits by summarizing key aspects in studies of agricultural, indus-
trial, and navigational benefits of water quality improvement. Section 3.6
briefly describes public water supply benefits, and Section 3.7 summarizes
the issues covered in the chapter. Case studies follow the text in each of
the appropriate sections. (The scenarios in Chapter 6 show how the case
studies can be applied in new situations.)
3.2 CATEGORIES OF BENEFITS: AN OVERVIEW
Since each household or firm undertakes different types of activities,
each is differently affected by water quality changes. A natural starting
point in appraising the various types of benefits of water quality programs is
to place them in broad classes. Figure 3-1 shows the categories of benefits
associated with water quality programs. The top part of the figure aligns
each benefit type with the uses made of water bodies. More perspective on
these benefit types is given by Table 3-1, which lists them according to
households or firms (the type of economic agent likely to receive them) and
the methods appropriate for measuring them. For example, the user benefits
category separates health benefits from recreation. Contingent valuation and
hedonic property value models are potential candidates for measuring both
3-1
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Potential
Water
Quality
Benefits
Current
User
Benefits
Intrinsic
Benefits
Direct
Use
Indirect
Use
Potential
Use
No
Use
1 — Recreational '- fishing, swimming, boating,
rafting, etc.
In Stream _— 1
1 — Commercial — fishing, navigation
1 — Municipal — drinking water, waste disposal
Withdrawal — Agricultural - irrigation
I— Industrial/Commercial — cooling, process treatment,
waste disposal, steam generation
i — Recreational'— hiking, picnicking, birdwatching,
photography, etc.
Near Stream _ Relaxation - viewing
— Aesthetic*- enhancement of adjoining site amenities
i — Near-term potential use
Option' 1
1 — Long-term potential use
• — Stewardship - maintaining a good environment for
everyone to enjoy (including future
Existence'^ family use-bequest)
' — Vicarious consumption — enjoyment from the
knowledge that others
are using the resource.
Figure 3-1. A spectrum of water .quality benefits.
categories of benefits. In addition, damage functions may be useful for
health benefits, while the travel cost and recreation participation survey ap-
proaches can measure recreation benefits.
Although they are not directly associated with use, option and existence
values are potential sources of economic benefits. Option value is the amount
that an individual may be willing to pay (over his expected user values) for
the right to use a water body--e.g., a river—in the future because uncer-
tainty exists either in the river's availability or in the individual's use of it.
Specifically, if an individual thinks he may want to use the river, but isn't
sure, then he may pay some amount each year for the right to use it. When
this payment exceeds the benefit the individual would receive from use, the
excess is the option value. Existence value is the willingness to pay simply
for the knowledge that a resource exists--!.e., the value an individual places
on a resource just because he knows it is there. Thus, because of a steward-
ship or related motive, an individual might be willing to pay something to main-
tain a river even though he knows he will not use it.
3-2
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Table 3-1. Classification of Benefit Categories
Type of
economic
agent
Benefit category
Type of measure-
ment method
Household: User
Recreation
Household: nonuser
Business and munici-
pality: user
Health
Option value
Existence value
Cost savings
Travel cost model
Contingent valuation survey
Recreation participation
survey
Hedonic property value model
Damage function
Contingent valuation survey
Hedonic property value model
Contingent valuation survey
Contingent valuation survey
Cost function
For businesses and municipalities, classification is more clearcut: The
effects on a firm's cost of production are the primary interest because the
role of water quality is exclusively reflected through these effects. Irriga-
tion, nagivatio'n, and .process uses are examples of water uses where cost
savings may arise.
Benefit classification should not be misinterpreted. In most'cases it is
impossible to separate all the sources of a benefit estimate. For example,
while willingness-to-pay estimates for a water quality improvement derived
from a hedonic property value model (discussed below) may be based on both
direct and indirect uses of the water body, the contributions of each cannot
be shown in practice. Equally important, overlaps should be expected be-
tween the methods used and the types of benefits derived. A contingent
valuation survey's (discussed below) estimate of willingness to pay for im-
proved water quality may include health, recreational, and nonuser benefits.
The exact composition will depend on the water body under study and the
nature of the questions used to elicit the information.
The classification scheme offered in this section simply shows possible
sources of benefits and may help identify measurement methods. However,
because this scheme does not fully define each of the methods or describe
how a water quality change is- introduced in each, the following sections offer
more detailed discussions of the various benefit estimation approaches, in-
cluding specific case studies that summarize previous work.
3-3
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
3.3 PRACTICAL CONCERNS: SELECTING A BENEFIT ESTIMATION METHOD
One of the first questions about benefit-cost assessment facing the prac-
titioner is how to select a method or set of methods for measuring the bene-
fits of water quality programs. This decision will be influenced by the fol-
lowing conditions:
The time and financial resources available.
The types of economic agents affected by the change in water
quality regulation.
The nature and magnitude of the changes in the water quality
regulations themselves.
Data availability.
These conditions largely determine the appropriate level of detail for a study.
In most benefit-cost assessments of water quality programs,.existing data and
results available from other studies will be sufficient. In complex cases, new
data and case-specific methods may be necessary. However, even in these
cases, when neither time nor resources are available, existing literature
and, to a lesser extent, ad hoc methods must be used.*
In practice, the ideal conditions routinely assumed in theoretical analyses
of benefit measures simply do not exist. Indeed, many benefit analyses of
environmental resources result in compromises that arise from a poor under-
standing of the exact association between water quality and particular activ-
ities of economic agents. However, while compromises may be necessary to
measure the benefits of water quality regulations for some water bodies,
every effort should be made to measure benefits based on willingness to pay
and cost savings—the only definitions for economic benefits that have clear
theoretical justifications.
Finally, because resources are limited, benefit-cost assessments of water
quality programs require that resources be wisely used—i.e., that they be
closely matched with the complexity of specific cases. Whether or not re-
sources are available, however, the practitioner must clearly understand the
features of each benefit measurement method to make an intelligent choice
among them.
*For example, while the Water Resources Council's current guidelines for
cost-benefit analysis, recommend use of the travel cost or contingent valua-
tion approach for estimating economic benefits of outdoor recreation services,
they acknowledge that practitioners may have to use ad hoc approximations
such as activity-day values—constant dollar values proposed for days of par-
ticular types of outdoor recreation (see Water Resources Council [1979])—
multiplied by projections of user-days.
3-4
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
3.4 HOUSEHOLD BENEFITS
There are several ways to measure the benefits of water quality pro-
grams. However, to compare the accuracy of* these measures, the practitioner
must also understand how individuals value different goods. Fortunately,
economics provides an objective way to measure these values using the basic
concept of a demand function.
Theory: The Demand Function
The first organizational guidepost that economics provides is the concept
of an individual demand function, shown in Figure 3-2. This function de-
scribes for any good, X, the maximum quantity of the good an individual
would be willing to purchase for each price of X. The downward slope of the
curve indicates that individuals are willing to buy more of X at lower prices
than at higher prices. The simple diagram in Figure 3-2 assumes all other
factors that might influence demand—including income, the prices of related
goods, etc.--do not change. Frequently, there is no need to actually meas-
ure a .demand curve. What it does is provide a basis from which the benefits
to households can be viewed.
time
Figure 3-2. The demand function and consumer surplus.
If the market process establishes a price at P , the individual will pur-
chase Q of X and make a total expenditure equal l8 P AQ O. Since the area
under tne demand curve measures the individual's maxfrnun? willingness to pay
for each unit of consumption, the total willingness to pay for Q is the entire
area—total expenditures plus the triangle P P.A. This difference—between
o j
3-5
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
what the individual actually pays and the amount he is willing to pay--is the
consumer surplus, or the dollar measure of the satisfaction an individual re-
ceives from consuming a good or service, less what he pays for it. As a dol-
lar measure of individual welfare, consumer surplus is not ideal, but most
studies have found it to be a good benchmark.
Practice: Methods for Measuring Household Benefits
Household benefits may be measured by using the travel cost method, a
contingent valuation survey, a recreation participation survey, the hedonic
property value method, or the damage function method. Advantages and dis-
advantages, data requirements, and key assumptions of each method are high-
lighted in this section. Case studies show how the methods have been used
in recent applications.
Travel Cost Method
One of the most popular approaches to describe demand for the services
of recreation facilities, the travel cost model,* has been used to estimate re-
creational benefits in a wide variety of applications. The -logic underlying
this model is simple. Recreators at a particular site pay an "implicit" price
for using the site's services through the travel and time costs associated with
visiting that site.f Since recreators visit a site from diverse origins, their
"travel behavior" can be used to analyze the demand for the site's services.
That is, all else being equal, any person will continue to travel to the site
until the marginal value of the last trip is exactly equal to its full costs
(i.e., the travel expenses and the opportunity cost of the time spent travel-
ing).
As a rule, the travel cost model estimates the demand for the representa-
tive individual. Therefore, to estimate the aggregate benefits of water qual-
ity improvement at a site, the practitioner needs to estimate how many recrea-
tors would use it. The solution to this problem will depend upon the data
used to estimate individual demand. For example, the visits made by resi-
dents of an origin zone, usually during a season, relative to the population of
that origin zone is .the quantity measure—a rate of use--conventibnally used
in the travel cost model. Since benefits are for the "representative" rate of
use, multiplying by the population of that origin zone will yield its aggregate
benefit estimates. Overall benefits would be the sum of the zone benefit esti-
mates .
*For further details see Dwyer, Kelly, and Bowes [1977], Freeman [1979],
Smith [1975], and Desvousges, Smith, and McGivney [1983].
fMost public recreation facilities either have no user fees or have nomi-
nal fees that do not reflect the marginal cost of a site's recreation services.
Thus, these fees are hot indicative of the equilibrium prices that would arise
if conventional market mechanisms allocated the services of recreation sites.
3-6
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Travel cost models also can be estimated by surveying users at a specific
recreation site. Then, the benefit estimates are for a "representative" recre-
ationist and estimates of total site use estimates, must be obtained independent-
ly. Possible sources of these estimates include the Corps of Engineers, the
Department of Interior, or parks and recreation departments at the State
level.
If a travel cost model can be estimated, it.can be used to estimate con-
sumer surplus. However, this is only part of the problem that must be
solved to appraise the benefits of water quality improvements. In addition,
the linkages between the regulations, the changes in water quality, and the
recreation decisions of individuals must be known. One possible linkage is a
change in demand for the services of a recreation site because individuals
wish to use them in one or more activities. A change in the level of water
quality may permit a wider range of uses, increase the individual's enjoyment
(and hence valuation) of existing uses, or both, which increases the demand
for the site's services at each (implicit) price.
Therefore, to evaluate the implications of a change in water quality for
an individual's economic well-being, water quality must be linked to the vari-
ables in a recreation demand function. Three ways for making this association
will be discussed in detail in Volume II. The case shown in Figure .3-3 offers
the most acceptable approach for linking water quality to recreation site de-
mand. It incorporates the effects of water quality as a determinant of the
demand in the travel cost model for a site's services (see Freeman [1979],
Chapter 8). Since little evidence generally exists on the variation in water
Implicit
Price E
$/visit
WQ«>WQ
Demand (WQ«)
Demand (WO)
0 Q1 Q2 Quantity of
Water Quality Treated as a Demand Change v'S't*/Y«»r
Figure 3-3. Illustration of the treatment of water
quality with travel cost demand model.
3-7
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
quality across the same water body, the travel cost model is difficult to imple-
ment empirically with a single site. Conceptually, however, a change in
water quality is assumed to shift site demand by providing a wider ranger of
activities. Thus, in Figure 3-3, the benefits associated with a change from
WQ to WQ* would be represented by the area HGFE. The scenario in Chap-
ter 6 shows an existing study that can be applied to new situations (see
Chapter 6, Section 6.4).
Data Needs, Key Assumptions/Limitations, and Features
The following checklist outlines the data needs and key assumptions and
features of the travel cost model.
Data Needs:
Origin—county of residence or zip code—for users of the
recreation site. These are often available from recreation
management agencies for samples of users.
Population size and summary measures for features of the popu-
lation in each origin zone (e.g., median family income, median
age, and median education). Sources include census data,
national and State recreation surveys, and site surveys.
Round-trip mileage from each origin to site. This information
can be calculated from maps.
Vehicle costs per mile and implicit time costs of travel. Travel
costs should be calculated as operating costs per mile for the
vehicle. Time costs can be estimated with the approximate
wage rate for the household head. One source is a wage and
occupation survey.
Key Assumptions/Limitations and Features:
The model is site specific. It measures the demand for the
services of a site, not total or general recreation demand.
The model measures only user benefits.
Consistency in the length of stay for each type of trip in ag-
gregate data. For example, all trips are treated as day visits
or as weekend visits.
A site's demand depends on its potential services for the re-
quired activities. (For example, a minimum-sized river seg-
.ment is necessary for power boating, while a river segment
with extensive locks and dams is not conducive to canoeing.)
The cost of time spent at the site is excluded. This suggests
that "full-cost" may not be expressed in a demand relation-
ship.
3-8
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
There are no good substitute sites available. If many substi-
tutes are available, the simple model will overstate the demand
for the site.
The travel cost is assumed to capture all the factors that in-
fluence the decision to recreate at the site. (For example,
this assumption implies that no changes in access, docks, or
other site features occur.)
The only purpose of the trip is to recreate at the site. If
this is not the case, the cost of the trip has a joint cost and
benefits are overestimated.
* •
CASE STUDY: ALTERNATIVE APPROACHES FOR ESTIMATING
RECREATION AND .RELATED BENEFITS OF
THE MONONGAHELA RIVER*
I ntroduction
The travel cost model for the Monongahela River study assumes that site
features or attributes affect the individual's ability to participate in recrea-
tional activities at any particular site, as well as the quality of the recrea-
tional .activities undertaken. It considers the demand for a recreation site as
a derived demand. That is, a site's services are desired because of the
recreational activities that can be undertaken at that site. Common sense
suggests that a recreation site's features or attributes will influence the de-
mand for its services. Since the level of water quality is a site attribute, a
basis is established for relating water changes to shifts in demand for a
recreation site's services.
Approach
The measurement approach examined numerous water-based recreation
sites from the Federal Estate Survey component of the 1977 National Outdoor
Recreation Survey. This survey provides specific information on the sample
recreationist patterns of use during a single season for each site. The
sample sizes for each site ranged from approximately 30 to several hundred
respondents and included information on individuals' recreation behavior,
socioeconomic characteristics, travel time necessary to reach the site, residen-
tial location/ and a variety of other factors. This information permitted the
estimation of individual travel cost demand models for each of the recreation.
sites.
x
Several advantages of this travel-cost model include:
*This discussion is taken from Desvousges, Smith, and McGivney [1983].
3-9
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Deriving individual " estimates for the time associated with
traveling to the site as well as the roundtrip distance for each
trip.
Using the opportunity cost of time to evaluate travel time and
estimating opportunity cost for each individual based on his
characteristics, including age, education, race, sex, and occu-
pation.
Considering for each site the potential effects of individuals'
differences in onsite time per visit.
A total of 22 individual site demand functions were estimated based on
this survey. For example, Equation (3.1) is a general description of one of
these site demand models:
In V ~ a + bTC + cY,
(3.1)
where
In V = the natural -log of the number of visits by a household to
the site in a recreation season.
TC = the travel cost per visit to the site, including out-of-pocket
vehicle operating costs and the opportunity cost of the time
spent traveling.
Y = family income.
The basic hypothesis of this study is that variation in the estimates of
a, b, and c across sites reflects the .effects of those sites' characteristics on
the representative individual's demand for each site's services. Thus, each
estimate provides the basis for describing how a change in any attribute
would affect demand.
The second step in the study involved estimating the relationship be-
tween variations in the .site-specific estimates of a, b, and c and each site's
attributes. The site characteristic information was obtained from records of
the U.S. Army Corps of Engineers and the water quality data from the U.S.
Geological Survey. Table 3-2 reports the site attributes, including the water
quality measures used in the model. Many other attributes, such as boat
launches, docks, and recreational facilities, were tried, but none was statis-
tically significant. For sites where the information on water quality was incom-
plete, the average value for all sites was used. This treatment of missing
values means that the estimated relationships will rely primarily on sites with
observed readings for the water quality variables.
Since precision in the estimates of demand parameters a, b, and c in
Equation (3.1) varied, a statistical procedure was used to account for the
quality of the estimates. Table 3-3 shows the estimated equations for the
demand parameters. As Table 3-3 shows, many of the attributes have statis-
tically significant effects on the demand parameters, particularly on the travel
3-10
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Table 3-2. Site Attributes Considered in Monongahela River Study
Variable
name
Description
Source
SHMILEa
ACCESS8
ARSIZE3
D0a'b
Number of miles of shoreline for the site
Number of developed multipurpose recrea-
tional areas plus the number of developed
access areas on the site
The ratio of the pool surface acreage
during the peak visiting period relative
to the total project area in acres
Dissolved oxygen based on monthly
readings
U.S. Army Corps
of Engineers
U.S. Army Corps
of Engineers
U.S. Army Corps
of Engineers
U.S. Geological
Survey
These variables were considered as monthly readings and as 4-month
averages in the specification of the demand parameter models.
K
DOM and DOV correspond to the average value of dissolved oxygen over the
four monthly observations and the variance about that average, respectively.
Table 3-3. Estimated Equations for Site Demand Parameter Estimates*
Variable
Intercept term
SHMILE
ACCESS
ARSIZE
Average DO
Variance in DO
A
a
1.51
(4.08)
0.0003
(1.25)
-0.0059
(-1.50)
-0.395
(-1.75)
0.0045
(1.07)
0.0005
(1.86)
Site demand parameter
6
-0.0246
(-9.48)
-0.00001
(-6.76)
0.00008
(2.81)
0.0033
(2.27)
0.00018
(5.99)
0.00001
(4.08)
estmates
c
0.000005
(0.308)
9.74 x 10~10
(0.09)
4.69 x 10~7
(2.56)
-1.94 x 10~6
(-0.18)
-1.22 x 10"7
(-0.60)
9.39 x 10"11
(0.01)
a
The numbers in parentheses below the estimated coefficients are the asymp-
totic (approximate) t-ratios for the null hypothesis of no association—the
larger the number, the more likely the hull hypothesis is rejected. These
equations show how the parameters of the individual site demand equations
vary with changes in the site's attributes.
3-11
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
cost parameter. The water quality attribute has a highly significant effect oh
travel costs, with the most plausible results obtained using the mean value of
dissolved oxygen over the 4-month summer period (June through September)
and the variance in dissolved oxygen about that mean. However, the small
variation in the water quality measures over the sample suggests these find-
ings be interpreted cautiously.
The model was used to evaluate the benefits of a water quality improve-
ment for users of the Monongahela River in Pennsylvania. This site was not
included in the 22. used to estimate the model. The model was applied to an
independent data set based on a household survey of residents in the Monon-
gahela River Basin (see the case study under Contingent Valuation below for
more details). The survey reported sufficient information on the respondents'
socioeconbmic characteristics, as well as specific portion(s) of the river used,
to construct individual demand curves varying by river site (at a total of 13
different sites) and by individual. The benefit calculations were as follows:
Estimates, of consumer surplus loss per user if the river were
no longer available for its current use—recreational boating.
• ' Estimates of the increment to consumer surplus associated with
• improving water quality from the current level that permits
boating, to a .level that would accommodate recreational fishing.
Estimates of the increment to consumer surplus associated with
improving water quality from the current level (boating) to a
level that would accommodate swimming.
The levels of dissolved oxygen used in the .benefit calculations for each
of these use designations were the values selected by Vaughan in Mitchell and
Carson [1982] in a water quality ladder developed for Resources for the
Future (RFF). The variance in dissolved oxygen was held.constant at levels
corresponding to those generally observed at the 22 sites. Table 3-4 provides
the RFF ladder thresholds for each activity.
Table 3-4. Specifications for the Dissolved Oxygen (DO)
Levels Associated With Use Designations
Use
designation
DO level
(percent saturation)
Beatable
Fishable .
Swimmable
Drinkable
45
64
83
90
These thresholds correspond to those used in RFF's
water quality ladder.
3These use designations were considered for benefit
analyses.
3-12
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Benefit Estimation
Figure 3-4 illustrates the model. DCWQt) corresponds to the representa-
tive individual's demand for the site's services at a water quality level of
WQj. Since the demand function is specified as semi-log in quantity, there is
no maximum price at which visits to the site will be zero. Therefore, the
benefit calculation required a maximum feasible price--P*. This was taken to
correspond to the largest travel cost incurred by any of the users of the
Monongahela River ($22.65 per roundtrip).
$/viiit
P* - $22.65
D(WQ2)
In visit
yr
Figure 3^4.General travel^ost demand-model lor a water
quality improvement.
The first benefit calculation involves the computation of the baseline area
P.ABP*—the loss in consumer surplus if the .site could no longer provide the
services available with a water quality of WQj to the user with a travel cost
of P.. Incremental benefit estimates were derived by estimating the addition
to consumer surplus associated with the increment to water quality. In Fig-
ure 3-4, a change from WQX to WQ2 would be shown as leading to an incre-
mental benefit of ACDB for the user at a travel cost of P.. Table 3-5 pro-
vides a summary of the average benefit estimate for each change and the
range of estimates over the survey respondents.
This case study provides a detailed model for incorporating site attri-
butes into estimating the demand for water-based recreation sites and, in
turn, for evaluating the benefits from changing one or more of those attri-
butes. Since It was developed from data primarily on flat-water recreation
sites, the model can be used to predict benefits for changes in attributes for
a sizable range of recreation sites. However, the actual estimates of the
benefits of water quality improvements in the study must be regarded as
tentative because of the limited available information on water quality. The
approach illustrated by this case study could be used in a wide variety of
applications. For one example, see Chapter 6, Section 6.4;
3-13
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Table 3-5: Benefit Estimates .from Generalized Travel
Cost Model With the Monongahela Survey Respondents
Benefits'
Use designations
Mean
Range
Benefits lost as a result of the loss
of ability to undertake boating
activities
Incremental benefits gained as a
result of water quality improvement
changing use designation from
boatable to fishable
Incremental benefits gained as a
result of water quality improvement
changing use designation from
boatable to swimmable
$53.35
$ 4.52
$ 9.49
0 to $70.80
0 to $8.60
0 to $18.30
The benefits are measured as the consumer surplus per user for the use of
the river during a single recreational season.
Survey—Contingent Valuation
The contingent valuation survey approach for estimating the benefits- as-
sociated with a nonmarketed commodity such as water quality improvements
involves asking individuals about their willingness to pay for different levels
of the commodity involved. Use of the survey approach requires that the
practitioner determine the aspects of changes in environmental quality individ-
uals value and convey these aspects to the respondent. The approach as-
sumes that individuals will accurately reveal their valuation of potential be-
havioral responses in hypothetical market experiments. These experiments
depend on a survey procedure and a survey instrument. The survey proce-
dure determines the appropriate sampling plan and specifies the general re-
quirements of the survey instrument. The survey instrument is the question-
naire used to elicit the respondents' answers.
A survey instrument is the cornerstone of the hypothetical market used
in the contingent valuation survey approach. It will generally consist of the
following sections [Rowe and Chestnut, 1981.]:
Introduction and statement of purpose
Nonvaluation questions
3-14
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Scenario development and market definition
Bidding or valuation questions.
The first two sections are self explanatory.and will not be discussed further.
The scenario .development and market definition section is considered the most
critical aspect of survey instruments because it must carefully present the
alternative levels of environmental quality. In the case of water quality,
scenario development describes the linkages between the regulatory action and
the resulting.change in water quality. Verbal or written descriptions, supple-
mented by visual props, are used in this activity. Scenario development must
be informative and realistic. It must portray the probabilities of the effects,
as well as the effects themselves.
After the hypothetical scenario is developed and the market is defined,
the bidding or valuation of the environmental commodity takes place. Several
questioning formats can be used:
Pi rect Question - - The interviewer directly asks the individual's
willingness to pay for a specified change in the amount of a
commodity—water quality--that has been carefully defined. No
cards or other aids are used to obtain the amounts.
Bidding Game—The interviewer defines the change to be evalu-
ated, suggests to the individual an amount representing the
value of the change (the starting point), and asks whether he
would .be willing to pay that amount. Based on the response,
the interviewer raises or lowers the suggested value by a
fixed amount and repeats the process until the individual
agrees no further change is necessary.
* Payment Card—This approach also does not directly ask about
willingness to pay. The interviewer explains the specified
change to be evaluated, provides the individual with a card
displaying an array of potential values, and asks him to select
a value or give any value for willingness to pay. These num-
bers range from zero to values judged to be outside the range
.of responses. Some surveys (notably Mitchell and Carson
[1981]) have adjusted the upper bounds of values on the cards
for higher income respondents. In addition, in an anchored
payment card format, some responses have been identified as
reflecting the share of an individual's taxes associated with
specific public programs, such as education and defense (see
Mitchell and Carson [1981])-
Bidding Game With Budget Constraint—This approach is a very
recent innovation for the bidding game format discussed in
Brookshire et al. [1982]. Before requesting a bid in the for-
mat explained above, the interviewer asks the individual to
estimate his after-tax monthly income and allocate it into ex-
penditure categories--for example, electricity, shelter, enter-
3-15
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
•.
I
I
tainrrient, savings, and a residual. After this information is
acquired, the interviewer conducts a bidding game with an
additional question: "Which of the categories of expenditures
would be reduced in order to make the proposed payment?"
Ranked Choice and Willingness to Pay--ln this approach, intro-
- duced by Rae [1981a, T981b], the interviewer provides individ-
uals with different hypothetical market outcomes—proposed pay-
ments and a specified level or change in water quality, for
example, to be ranked. These ranks are then used in a statis-
tical analysis to estimate the individual's willingness to pay.
The question format and description of the hypothetical market are important
determinants of the quality of the estimates derived from a contingent valua-
tion experiment. The results of a comparative analysis of the direct question,
payment card, and bidding game formats in Desvousges, Smith, and McGivney
[1983] suggest that, for questions associated with water quality, the question
format has some effect on the average willingness to pay. The starting point
for the bidding game appears to influence the average values.
Two basic types of biases can arise in designing the format of a contin-
gent valuation study. Since Schulze, d'Arge, and Brookshire [1981], Rowe
and Chestnut [1981], and Mitchell and Carson [1981] have all discussed these
biases in detail, a brief overview of their conclusions is provided in Table 3-6,
which defines the bias, identifies the studies that considered its potential
effects, and summarizes the current understanding of its effects.
Overall, the results seem to suggest that starting point bias may be the
most important consideration (aside from the hypothetical nature of the ques-
tions, which has not received sufficient testing to fully gauge its implications)
in using the contingent valuation framework. Most of the other potential
sources of bias can be controlled in the structuring of the instrument and the
explanations provided to sample respondents. Several additional technical
assumptions are highlighted below in a summary of data requirements and key
assumptions of the approach.
Data Needs, Key Assumptions/Limitations, and Features
The following checklist outlines the data needs for the contingent valua-
tion survey approach, along with its key assumptions and features.
Data Needs:
Survey of individuals designed to be representative of affected
population.
Clearly defined and pretested survey instrument. In-person
interviews are generally more reliable than telephone or mail
surveys.
3-16
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Table 3-6. Summary of Biases in Contingent Valuation Experiments
Type of bias
Definition
Studies that have
tested for bias
Summary of
current results
General
Hypothetical
Error introduced by posing hypothetical
conditions rather than actual condi-
tions to an individual; response may
not be a good guide to actual actions
individual would take
One known test--
Bishop-Heberlin
[1979]
Strategic
Attempt by respondents to influence out-
come of study by systematically over-
or under-bidding so action favors
their true interests; strategic
responses depend on how payment scheme
is defined and whether it is believed
At least eight tests
(see Schulze,
d'Arge, and
Brookshire [1981]
for summary;
Cronin [1982])
Instrument
Related
Starting
point
Contingent valuation experiments using
bidding game format have started with
suggested payment and use yes or no
responses to derive final willingness
to pay; suggestion may be perceived as
appropriate bid
At least five tests
(see Schulze,
d'Arge, and
Brookshire [1981]
and Rowe and
Chestnut [1981])
Vehicle
Information
Interviewer
Characteristics of proposed mechanism •
for obtaining respondent's willingness
to pay may influence responses
Effect of information provided to
respondent on costs of action under
study or other dimensions of problem
may affect responses
Responses vary systematically according
to interviewer
At least four tests
(see Schulze,
d'Arge, and
Brookshire [1981]
and Mitchell and
Carson [1982])
At least four tests
(see Schulze,
d'Arge, and
Brookshire [1981]
and Mitchell and
Carson [1981])
Two tests--
Desvousges, Smith,
and McGivney
{1983] and
Cronin [1982])
Some indication that
hypothetical nature
of question did
influence responses,
but could not dis-
tinguish this effect
from instrument-
related biases
Very little evidence
of strategic bias
except for Cronin
[1982]
Some differences in
opinion over impor-
tance of starting
point bias;
Mitchell-Carson
feel starting point
bias is important,
and Desvousges,
Smith, and McGivney
[1983] provide some
support; Schulze,
d'Arge, and
Brookshire [1981]
feel it is more
limited
Some evidence of
effects in at
least two studies
Limited evidence of
effects
No evidence of bias
Bias present
The definitions and results summarized in this table are based on Schulze, d'Arge, and Brookshire [1981],
Rowe and Chestnut [1981], and Mitchell and Carson [1981].
3-17
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Key Assumptions/Limitations and Features:
Individuals' responses to hypothetical questions are assumed to
be indicative of their actual valuations of the changes de-
scribed in the questions.
Careful tests are required to determine starting point effects,
appropriate mechanisms for payment, and consistency of re-
sponses with other budgetary requirements.
Careful Control is required over information given respondents
so answers are based on the same information in each inter-
view.
****************
CASE STUDY: RECREATION AND RELATED BENEFITS
OF WATER QUALITY IMPROVEMENTS OF THE
MONONGAHELA RIVER*
Introduction
This contingent valuation survey measured the recreation and related
benefits of water quality improvements in the Monongahela River Basin in
Pennsylvania. It compared alternative question formats for asking individuals'
willingness to .pay and measured both user and nonuser values. In a house-
hold survey conducted by 9 professional interviewers from the five-county
area, an 80 percent response rate was obtained from a clustered random
sample of 393 households.
Approach
In any contingent valuation study, the survey questionnaire is the key
element for providing plausible results. By dividing the questionnaire into a
version for each question format and distributing each version equally among
the interviewers, the Monongahela study compared the techniques,
A water quality ladder, developed by RFF (see the travel cost case
study for more detail), was used to establish a linkage between levels of
water quality and the associated uses for recreation. Tied to scientific meas-
ures of water quality, the ladder steps permit the respondent to give his will-
ingness to pay for the various levels of water quality.
The contingent valuation method requires a way to make the hypothetical
payment for water quality improvements. User fees, increases in sales taxes,
and increases in water bills are among the alternatives used. This study
expressed the additional annual amounts as taxes and higher consumer prices.
Also used by Mitchell and Carson [1981], this method corresponds roughly
with how a respondent actually pays for water quality improvements.
*This discussion is taken from Desvousges, Smith, and McGivney [1983].
3-18
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Each version of the questionnaire used a different technique to elicit the
respondent's willingness to pay. The iterative bidding technique was used in
two versions, with the interviewer starting the bidding at $25 in one version
and at $125 in another. The direct question techniques and the payment card
were also used. The payment card contained values arrayed from $5 to $775,
but no other information.
All versions of the questionnaire required the respondent to place a
value on a degradation in water quality in the Monongahela from its present
overall level of beatable to a level where the river was unsuitable for any
recreation. Additional amounts were elicited for water quality improvements
to support fishing and swimming. These amounts reflected actual use and
potential use in the future, with a followup question requiring the respondent
to break down the amounts into actual and potential use.
. Benefit Estimation
For each version of the questionnaire, Table 3-7 presents the average
amounts users and nbnusers of the Monongahela River were willing to pay for
Table 3-7. Willingness to Pay for Three Levels of Water Quality9
($/yr)
Users
Nonusers
Combined
Mean
Standard deviation
Number of respondents
Payment card
117.9 (47.1) 82.8
117.0 (53.8) 104.7
17 37
93.8 (71.6)
108.9 (92.8)
54
Mean
Standard deviation
Number of respondents
Direct question
98.2 (47.4) 34.5
103.5 (81.5) 66.4
17 34
55.7 (38.8)
85.2 (71.8)
51
Bidding Game: $25 starting point
Mean 59.5 (42.4) 51.4
Standard deviation 38.1 (31.9) 53.1
Number of respondents 19 39
54.1 (48.4)
48.5 (47.1)
58
Mean
Standard deviation
Number
Bidding Game: $125 starting point
194.4 (109.4) 79.2
136.5 (129.2) 102.5
16 32
117.6 ( 89.3)
126.0 (111.6)
48
As defined in Section 3.2, numbers in parentheses are individuals' estimated
mean option values and corresponding standard deviations.
3-19
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
avoiding a decrease in water quality and for raising it to swimmable levels.*
Several conclusions can be drawn from the results in Table 3-7. The esti-
mates of willingness to pay—regardless of method used to elicit the amount—
are quite substantial, ranging from $35 to $195 per year. The results are
sensitive to the method used; the payment card with the $125 starting point
bidding game gave higher amounts than the direct question with the $25 start-
ing point bidding game, there is some indication of a starting point bias,
but the evidence is not conclusive.
Users of the Monongahela generally were willing to pay higher amounts
than nonusers, with their average values ranging from $59 to $194 compared
to a nonuser range of $34 to $83. Estimates of option value are about half
the user's willingness-to-pay values and are statistically significant for both
users and nonusers. The results imply that benefit estimates based solely on
recreation use may substantially understate the total benefits of water quality
improvements. The approach illustrated by this case study could be used in
a wide variety of applications. .For one example, see Larson [1981].
Survey--Recreatioh Participation Models
Many State and Federal agencies undertake surveys of the general popu-
lation in an effort to identify household participation patterns for recreational
activities. As a rule, these surveys provide detailed information on house-
hold characteristics and on the types and amounts of participation in outdoor
recreation. These surveys have been used to .estimate recreation participation
models. Such models are neither demand nor supply relationships but sum-
maries of all the determinants of the likelihood that an individual will parti-
cipate in recreational activities—for example, boating, fishing, or swimming--
as well as of the level'of participation in these activities. Generally, these
models divide the participation decision into two steps: determining whether
a person participates in a particular activity and modeling the expected num-
ber of days (or trips) he spends at the.activity over a season.
These models have been developed from a framework that views the indi-
vidual as maximizing well-being by selecting levels of service flows for his
consumption. Individuals produce these service flows by using time and/or
purchased goods and services. For example, the level of participation in a
recreation activity is one measure of recreation service flow that requires the
person's time, any equipment associated with the activity, and the services of
a recreation site as inputs. Participation models, which describe the final
result of the activity, are influenced by each. element in the production of
service flows.
*These mean amounts are calculated exclusive of the respondents who
rejected the approach and those who were shown to be outliers by a statis-
tical analysis. For a complete discussion of the procedures used to make
these determinations and the small differences that result from the exclusions,
see Desvousges, Smith, and McGivney [1983].
3-20
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
This perspective is important because it establishes a natural association
between participation and travel cost models. Travel cost models are demand
models for the services of a recreation site. These services add to the pro-
duction of the recreation service flows. In addition, benefits measured by a
participation model should, for consistency, use the demand for the recreation
service flow and not for the recreation site, but these demands are difficult
to isolate.* This often makes assignments of benefit estimates resulting from
a change in water quality arbitrary (see Davidson, Adams, and Seneca
[1966]).
Instead of identifying the recreation facilities used by the survey, re-
spondents, participation surveys have, as a rule, required crude measures of
recreation supply characteristics to be merged. In principle, this merging
permits the use of summary measures of water quality for regional areas as
determinants of the likely participation and the level of participation of the
representative household. Yielding crude approximations at best, this prac-
tice reflects the paucity of data in this area.
However water quality is introduced, it should be emphasized that the
results of these models are estimates of the levels of use of recreation activi-
ties and not economic benefits. Methods for measuring these values must be
developed independently.
Data Needs, Key Assumptions/Limitations, and Features
The following checklist outlines the data needs for the participation sur-
vey method, along with its key assumptions and features.
Data Needs:
Survey of recreation patterns of the genera! population, with
socioeconomic detail and identification of residential location
(preferably in more detail than State of residence).
Identification of sites used for recreation activities, or at least
some measure of the supply of recreation facilities, is highly
desirable. .
Measures of water quality for sites used by respondents, or
linkage between water quality and capacity-related measures
for recreational activities.
Key Assumptions/Limitations and Features:
An independent estimate is required of an individual's willing-
ness to pay for a day or a trip spent in each recreational
activity.
*See Deyak and Smith [1978] and Bockstael and McConnell [1981].
3-21
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
The demand and supply relationships are assumed to remain
stable.
Model specification (i.e., two-step partition of participation
decision and level of participation) is assumed to be correct,
and functional forms are assumed to be adequate approxima-
tions.
Measures should be provided at a general level, not on a site-
specific basis.
****************
CASE STUDY: A PARTICIPATION SURVEY APPROACH TO
VALUING WATER QUALITY IMPROVEMENTS* ,
Introduction and Basic Logic
The Vaughan-Russell [1982] study, the most ambitious and detailed
application of a recreation participation model to date, focused on the recrea-
tional fishing benefits that arise from a change in water quality. It used the
fact that more "desirable" freshwater sport fish--coldwater and certain warm-
water species—require better water quality. Improved water quality may
alter the types of fish that can be supported in a water body. Assuming the
supporting recreation facilities are available, Vaughan-Russell suggest that
there will be a change in the type of fish (and perhaps a net increase in the
level of fishing participation) from less desirable to the more desirable varie-
ties. The sources of benefits from the water quality change arise from:
The change in the composition of fishing activities
Any net increase in the level of .participation {in fishing.
to implement this logic on a national scale, the Vaughan-Russell objec-
tive, requires the following:
Measuring the availability of freshwater bodies for fishing and
their water qualities at a geographically disaggregated level.
Modeling and measuring the influence of water quality on par-
ticipation in recreational fishing and on fishing .activities by
type of fish sought.
Measuring the economic benefits according to the type of fish
sought.
*This study is taken from Vaughan and Russell [1982].
3-22
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
This review highlights some of the key elements in this application of partici-
pation models.
Approach
The first step in the analysis was to estimate the available "fishable"
waters. Using dissolved oxygen and suspended solids criteria, Vaughan-
Russell projected total fishable water on a pre- and post-policy basis. - For
each policy scenario, they used the following steps:
1. Calculate the percentage increase from baseline levels of total
fishable water as represented in the RFF water quality network
model.
2. Estimate the policy impact by applying the improvement factors
from Item 1 above to a national baseline of fishable acres per
capita.
3. Estimate the change in composition of fishing activities by
using the water quality network model to calculate fishable
water by species type.
These steps are based on a recreational fishing participation model that
divides a person's participation choices into three decisions:
1. "Decide whether or not to fish"
2. "Decide what to fish for"
3. "Decide on a level of participation."
The model implies that the amount of fishable water available affects the
probability of an individual's being a fisherman. Then, the suitability of
water quality to support a class of fish (e.g., coldwater, warmwater, game,
and rough) affects the type and level of fishing activity. Policies that
change water quality affect the availability of each type of fishable water. In
the Vaughn-Russell study, this relationship established the necessary
technical linkage between water quality and behavior (see Chapter 1).
The empirical analysis of participation was based on the 1975 National
Survey of Hunting, Fishing, and Wildlife Associated Recreation. For Parts
(1) and (2) of the fishing decision, probability models were estimated using
several statistical techniques. The models included a wide array of socio-
economic variables (e.g., age, sex, income, region, residency in metropolitan
area, residency in State with coastline, and total acres of fishable water in
the State per capita). The estimated effect of acres of fishable water per
capita on the likelihood of participation in fishing was positive and statisti-
cally significant.
U.S. EPA Headquarters Library
Mail code 3201
3-23 1200 Pennsylvania Avenue NW
Washington DC 20460
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Several methods were considered for estimating the second-stage proba-
bility models--!.e., for the particular type of fishing selected. The final set
of models for these probabilities was based on three overlapping fishing
choices--some trout, some bass, and some .-rough fish. The most significant
effect of the fishery-specific water quality variable .was the coldwater game-
fish category. Nonetheless, the signs of 'the effects of the water quality
variables generally agree with a^ priori expectations for all categories of
fishing.
In the last component of the model, the number of days spent fishing in
each category were estimated. The participation model requires the use of all
three components to evaluate the implications of a change in water quality on
the types of fishing chosen. Table 3-8 illustrates the results of one of the
Vaughan-Russell scenarios—adoption of Best Practicable Technology (BPT) for
the predicted changes in the mix of fishing .activities undertaken. The last
three lines in Table 3-8 provide the "bottom line" implications of the model in
physical terms. They are not benefit estimates but, rather, increases in the
number of fishing days of various types. If benefit estimates are to be de-
rived from the model, these fishing days must be valued.
Table 3-8. Vaughan-Russell Model--Predicted Effects
of BPT Regulations on Participation
Change from base case
Change in relevant variable BPT/BASE"
Probability of being a fisherman
Probability of doing some:
Trout fishing
Bass fishing
Rough fishing
Days per capita per year:
Trout
Bass
Rough
Total days per year:
Trout
Bass
Rough
+0.0001'
+0.0076
-0.0142
-0.0039
+0.02
+0.34
+0.51
+7-2 x 106
-1.3 x 106
+5.6 x 106
SOURCE: Vaughan and Russell [1982], Table 6-1. '
aThe logit estimates using a sample size of 5,000 were used for these esti-
mates .
This may seem an inconsequential change in the probability of being a
fisherman, but it implies an>increase of 20,000 fishermen per year.
3-24
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Benefit Estimation
To value these fishing days, Vaughan-Russell conducted a separate sur-
vey of fee fisheries in the United States and used the results of the survey
to estimate a travel cost model, accounting for the effects of a number of
characteristics of the individual fisheries.* Separate models were estimated
for trout and catfish (as a basis for valuing the rough fishing). The esti-
mates of consumer surplus were converted to a per day per person basis.
Considering the differential in these values between trout and catfish, the
estimates ranged from $1.77 to $8.06 in the final models. These estimates
provided the basis for valuing the increments to fishing days projected in the
participation model. When the relevant daily per capita consumer surplus is
applied to the estimated increments in fishing days, the incremental benefits
can be calculated from the Vaughan-Russell model.
One of the most important potential limitations to the Vaughan-Russell
methodology is the procedure used to value fishing days. The relevant bene-
fit measure is the demand for fishing as a recreational service flow, not the
measure estimated from the demand model for a site's services. While there is
a correspondence (see Anderson [1974] or Carlton [1979]), the relationship
between the two will depend on the nature of the other inputs to fishing and
the activities undertaken at the site. Since the Vaughan-Russell model relies
on a very specific definition of the recreational activity and treats trips to
the fisheries as single-day visits, the discrepancies may not be great. How-
ever, the transfer of the relationship between travel cost and participation
models implied by their framework may not be possible in other applications.
There should be little doubt, even with this cursory review, that the
Vaughan-Russell model represents an- enormous undertaking and is the best
effort available to date for modeling recreation participation. The approach
illustrated by this case study could be used in a wide variety of applications.
For one example, see Chapter 6, Section 6.4.
Hedonic Property Value Method
Two types of recent models use market data on either property values or
real wages along with quantitative measures of environmental amenities to esti-
mate individual willingness to pay for a change in one or more amenities. Re-"
searchers have applied both property and wage models to value air quality
but have used only property value models in the case of water quality.
These models, known as hedonic models, use two assumptions: (1) par-
ticipants in a market accurately perceive the characteristics of different hous-
*The characteristics did not include water quality because the fisheries
were separated by the type of fish.
3-25
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
ing sites—including water quality-and, in making their location decisions,
will consider them along with the prices for the housing units; and (2) there
exists a continuous array of combinations of these characteristics across dif-
ferent housing sites within the market. An ideal market process will ensure
that equilibrium housing prices and rents will reflect the marginal valuations
of the characteristics.
Data Needs, Key Assumptions/Limitations, and Features
Only a checklist on data needs, key assumptions, and features is pre-
sented because the resources required to use this method would exceed those
available in most States. A detailed discussion and a related case study will
be presented in Volume II.
Data Needs:
Property values (preferrably sale price) for residential sites
around water bodies with different water qualities in the same
housing market.
Information on other site and neighborhood characteristics that
may affect property values.
.Information on individuals' perceptions of water quality and
relationship to available physical measures of water quality.
Key Assumptions/Limitations and Features:
Market equilibrium
Full knowledge of the implications and effects of water quality
Ability to determine extent of market and specify relationships
for hedonic price and demand functions
Full adjustment and ease of mobility.
Damage Function Method
The damage function method applied to valuing the benefits from water
quality improvement is most relevant for the effects of water quality on human
health. In principle, this approach examines all the possible physical effects
of each type of emission into a water body. However, usually only the health
effects are considered.
To use this approach it is necessary to estimate, for each class of effect
(i.e., chronic versus acute), health impacts that stem from the relationship
between the physical effect and the concentration of the relevant water pol-
lutant, as well as any other factors that might influence the pollutant's
impact. These relationships are the damage functions. They are used to
estimate the physical effects of specified changes in water quality as measured
3-26
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
by the concentrations of each of the individual pollutants. The method does
not provide a way to value the physical effects so that independent estimates
of the health benefits must be developed.
Data Needs, Key Assumptions/Limitations, and Features
The following checklist outlines the data needs for the damage function
method, along with its key assumptions and features. A detailed case study
will be presented in Volume II.
Data Needs:
Measures of concentration of relevant pollutants in water used
by population over time and over .the geographic location.
Measures of features of population and health patterns.
Measures of other exogenous factors that may also affect ob-
served health patterns of population.
Key Assumptions/Limitations and Features:
Provides largely statistical summaries of data on existing popu-
lation experiences.
Assumes no behavioral substitition on the part of populations
in response to levels of each pollutant. :
Produces results sensitive to the statistical procedures used to
estimate the models.
Maintains the primary advantage allowing classification of
effects according to physical impacts.
3.5 BUSINESS BENEFITS
Benefits from the water quality programs can accrue to firms as well as
to households because many provide for a wide spectrum of uses for rivers
and streams, including industrial/commercial, agricultural, navigation, and
municipal water supply uses. Measurement of business benefits are often
easier because market prices are usually available to value these benefits.
Theory: The Supply Function*
In addition to the demand function, discussed earlier, economics provides
a second organizational guidepost for measuring benefits—the supply function,
*This discussion is a summary of the discussion in Just, Hueth, and
Schmitz [1982]/Chapter 4.
3-27
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
shown in Figure 3-5. The supply curve shows the maximum quantity of out-
put of good X the firm is willing to supply at each relevant price. If the
market establishes the price at P , the firm will produce the quantity Q .
The upward slope of the curve in Tigure 3-5 indicates that the firm is willing
to sell more at higher prices than at lower prices, assuming that factors in-
fluencing the supply function—the prices of inputs, such as labor, energy,
machinery, and technological improvements—do not change.
Price
$/X
Quantity X
time
Figure 3-5. Supply function and the producer surplus.
The concept of producer surplus is used as the general measure for a
change in the welfare of a firm. In Figure 3-5, producer surplus is shown
as an area equal to the area above the supply curve and below the price line
for the firm or industry. Whether a producer is better off can be determined
by examining the change in producer surplus. Producer surplus provides a
measure of a change in welfare for a firm because its welfare is measured
directly in dollars of cost savings. This view of a firm is a simplified model
that does not include important differences among firms or the distribution of
profits among owners and resource suppliers.
Practice: Cost Savings Method for Measuring Business .Benefits
The estimation of a firm's benefits from water quality has been much less
sophisticated than the estimation of household benefits. The primary focus
has been on estimation of the cost savings associated with the water quality
change. The estimates are derived largely from engineering cost estimates.
In principle, economic cost functions could provide the basis for these esti-
mates, but, in practice, they have not.
The hypothetical example for business benefits is for irrigation, but the
same method could be adapted to the other situations. The incremental pro-
ducer benefits that might arise from water quality programs are:
3-28
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
• Reductions in industrial cost—firms using water in processes
may have reduced production costs because either less treat-
ment of water is required or less maintenance is required for
pumps, pipes, and other parts of the production process.
Reductions in agricultural costs—farmers using water for irri-
gation may have reduced production costs because less mainte-
nance of irrigation equipment is required or less fertilizer per
bushel of crop is needed.
Reductions in navigation costs—barges and other water trans-
port conveyances may have reduced risk of accident. Ship
maintenance may be reduced.
From .these simple examples two major points arise. The first is
that the focus on incremental benefits of water quality decisions will
mean these benefits are considerably smaller than they would be if meas-
uring the total benefits of all water regulations were the primary objec-
tive of the benefit-cost assessment. That is, the baseline is important
in .measuring the benefits of a particular decision. The second point Is
that the relative orders of magnitude of these benefits will be very speci-
fic to the Individual water bodies evaluated. For example, the comple-
mentary attributes necessary for recreation, such as access and overall
surroundings, might be at very low levels in some instances, while the
potential for producer benefits from other designated uses is very large.
This point emphasizes the importance of the focus in .the proposed water
quality standards program on selecting key segments and considering each
on a specific basis.
Data Needs, Key Assumptions/Limitations, and Features
The following checklist outlines the data needs of the cost savings
approach, along with its key assumptions and features.
Data Needs:
Cost data for firm (see Chapter 4 for details)
Demand information such as market prices and responsiveness
of sales to price changes.
Key Assumptions/Limitations and Features:
All products and inputs (labor, machines) are bought and sold
In markets that are perfectly competitive; that is, no buyer or
seller has influence over market prices.
The supply curve reflects the marginal social cost of producing
the product or service. This implies that neither external
costs nor subsidies are present in the market. (This is un-
likely for the irrigation example because of the various legal
and regulatory influences in the market, but it is a useful as-
sumption to simplify exposition.)
3-29
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
A shift in the supply of a :producer's services will not affect
the price at which they are sold in the market.
* .
*
CASE STUDY: IRRIGATION. BENEFITS
Suppose that a State is considering a change in the designated uses.of a
river segment to provide for fish and wildlife propagation. As a byproduct
of this use designation, the quality of water available for irrigation is as-
sumed to improve, thus shifting the supply of irrigation services—because
more high-quality water is available for irrigation—outward from Sj to S2, as
shown in Figure 3-6.* In an application, the practitioner may have limited
data on some costs, but seldom enough to estimate the entire supply function.
By supplementing the available data (e.g., from the Bureau of Reclamation or
the U.S. Department of Agriculture [USDA]) with assumptions based on com-
mon sense, a rough cut at the problem can be obtained. For example pur-
poses, the entire supply curve is drawn.
$/103 gat
US
o>
1
<*•»
o
I 1
a.
Change in
~ Producer Surplus
irrigation
services
100 200
Quantity of Irrigation Service
Figure 3-6. Irrigation benefits.
300 103 gal/yr
*The stream's flow is assumed to be strong enough that the increases in
irrigation will not noticeably reduce it. Prior to the change in use designa-
tion, the limiting factor is assumed to be water quality rather than flow.
3-30
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
The task is to calculate the irrigation benefits from this shift in supply--
the change in producer surplus attributed to the implementation of the new
designated use. The crosshatched portions of Areas a and b in Figure 3-6
are the two components of the change in producer surplus. Area a is the new
producer surplus resulting from reduced cost of the original volume of irriga-
tion services (100). Area b is the new producer surplus on the additional
volume of irrigation attributable to the reduced cost resulting from the water
quality standards.
As drawn, Area a is a parallelogram showing the cost savings on the
original irrigation volume. Using the formula for the area of a parallelogram:
.Area a = (side) x (perpendicular distance to parallel side)
Area a = $(2-1) x (100-0)
Area a = $(1) x (100)
Area a - $100/year.
Area b is the additional irrigation induced by the decrease in costs due
to the water quality standards. Using the formula for the area of a triangle:
Area b = ^(base) x (height)
Area b = ^$(2-1) x (300-100)
Area b= %$(D * (200)
Area b - $100/year.
Thus, the change in producer surplus—the measure of firm benefits—
attributable to the cost savings from the additional river uses in the water
quality standards is $200 a year.
There are several important caveats to the forgoing simple example for
estimating firm benefits:
The costs are not quite as simple as in this example. However,
the basic measurement concept still applies.
The assumptions required for the example are stringent ones,
but they do provide a workable approximation for many indi-
vidual river segments.
In the cases of agriculture and nagivation, institutional factors
in those markets may distort the true social cost. For ex-
ample, the subsidization of waterway activities and the regu-
lated rates in railway and highway transportation may violate
the assumptions of perfect competition in those markets.
3-31
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
If the assumption of perfect competition in the input markets
does not hold, the producer surplus may actually accrue to
providers of labor or capital services.
Market power resulting in control over market prices and out-
puts will distort the supply relationships and make prices
higher than in competition.
3.6 PUBLIC WATER SUPPLY BENEFITS
Reductions in treatment requirements for municipal water supplies consti-
tute another potential source of benefits from water quality program
decisions. By having a use designation that provides for fish and wildlife
propagation, a city that uses a river as a water supply source may be able to
provide water with less treatment than if the use designation were agricul-
tural or industrial. Once again, it is essential to remember that the focus of
the benefit measurement .should be the incremental benefits attributable to the
particular policy, not total benefits from all water regulations.
Since calculation of these benefits could proceed exactly as in the case
.of business .benefits, no case study is provided. The same key assumptions
apply as with business benefits.
The critical issue of toxic substances or toxic pollutants, which would
apply to a public water supply, is not considered in this handbook. As more
information becomes available on the extent and effects of toxics and their
relationships to the water quality regulations, a change in this focus may be
warranted. This is an issue that bears future scrutiny because of .potential
health benefits from reducing toxic pollutants.
3.7 SUMMARY
This chapter has reviewed basic benefits concepts and the approaches
used to measure them. An individual's willingness to pay is the centraKtenet
underlying all the methods discussed in this chapter. Even though all assess-
ments may not require the practitioner to use measurement methods, willing-
ness to pay provides an organizing principle for even qualitative assessments
of benefits.
Major points developed in the chapter include:
The assessment should be tailored to balance the complexity
and importance of the policy action to the available resources.
• The change in consumer surplus should be used as the meas-
ure of willingness to pay for improvements in the well-being of
households.
3-32
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
The travel cost method can provide willingness-to-pay-based
measures of water quality changes under many conditions. It
measures only user benefits and is sensitive to the treatment
of time costs and substitute sites. The travel cost method
estimates demand for site services.
•The contingent valuation survey method also provides wifling-
ness-to-pay-based measures of water quality changes under
many conditions. It can measure both user and intrinsic bene-
fits and is sensitive to questionnaire design and administra-
tion.
The participation survey method provides measures of changes
in level of use—visitor days—for a recreation activity and
often requires ad hoc valuation of use to develop benefits.
Cost savings can provide estimates of willingness-to-pay-based
measures (producer surplus) of changes in the economic well-
being of firms.
3-33
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
CHAPTER 4
MEASURING THE COSTS OF WATER QUALITY PROGRAMS
4.1 INTRODUCTION
Why is opportunity cost the basis for measuring costs? What types of
approaches are available to measure costs? What practical problems arise in
measuring costs? Are there any examples of cost measurement?
This chapter focuses on determining the incremental costs of a water
quality program. It defines various cost categories, discusses their relation-
ships, and reviews costing methods relevant for users of this handbook. It
covers costs for both industry and publicly owned treatment works (POTWs)
and presents several case studies to illustrate how the general principles are
applied. Although the examples are oriented toward potential water quality
standards decisions, they are.general enough for use.in other applications.
Throughout this chapter, opportunity costs—the value to society as a
whole of a resource's best alternative use—provide the measurement basis for
costing in a benefit-cost assessment. Engineering and accounting cost esti-
mates may differ from opportunity costs because of cost-sharing .mechanisms
such as taxes and subsidies. In these cases, the practitioner faces the dif-
ficult task of determining the value of opportunity costs.
This chapter suggests practical approaches to costing and highlights
some of the more difficult issues, which will get .more detailed treatment in
Volume II. Specifically, Section 4.2 presents the basics underlying the meas-
urement of costs, and Section 4.3 describes two general approaches, engineer-
ing and econometric, to measuring costs. Section 4.4-defines types of cost
and major cost categories to be used in an assessment. Section 4.5 discusses
the practical aspects of determining costs—including data, sources, the use of
indexes, and major factors influencing cost estimates—and concludes with a
sample data form. Section 4.6 describes the engineering methods for estimat-
ing costs, and Section 4.7 provides examples using the engineering methods.
Section 4.8 offers some general cautions for measuring costs. Finally, Section
4.9 summarizes the chapter's major points.
4.2 MEASURING COSTS: THE BASIC CONCEPTS
This section reviews the fundamental economic principle of cost: oppor-
tunity cost, which measures the cost of any resource in terms of its next
best alternative use. That is, the value of forgone alternative uses for any
resource provides the basis for estimating the cost of any specific use of that
resource.
4-.1
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
For a water quality decision, opportunity costs include both explicit
costs (e.g., wages and salaries, or payments for materials and energy) as
well as implicit costs (what self-owned and employed resources could have
earned in their best alternative uses). For example, the implicit cost of labor
is the highest wage the owner could receive for his labor services.
For firms and households, the opportunity costs are the private costs of
a regulatory action. If the action also negatively affects others, either
households or firms, additional costs are incurred—in technical terms, exter-
nal costs. In a benefit-cost .assessment, the opportunity cost to society is
the relevant measure of cost, the sum of private and external costs.
The economic guideposts of supply and demand functions described in
Chapter 3 can be used to view the opportunity cost concept. These functions
are combined in Figure 4-1, which illustrates a market for good X. The de-
mand curve, D, shows the amount demanders are willing to buy at each of
several prices, while the supply curve, S, reveals the amounts suppliers will
provide at various prices. Market forces'will cause the price to settle at P0,
with the resulting quantity at Q0.
Price
$/Ox
Deadweight
lots
Q! QQ Quantity/time
Figure 4-1. Measurement of opportunity/costs.
One way to view the cost of a regulatory action is to suggest it will im-
pose costs on firms, resulting in a shift of the industry supply curve from S
to S1. In this case, the regulation causes a shift in the curve by an amount
that recovers the costs of compliance—shown by the hatched area in Figure
4-1. These costs constitute one element in the opportunity cost of the regu-
lation—the additional opportunity costs (e.g., extra operation and mainte-
nance) required to meet the regulatory standard. The second element of op-
portunity cost is shown by the shaded triangle—the loss to society because
the produced and purchased quantity of X is reduced from Q0 to Qt when
supply shifts. In technical terms, some producer and consumer surplus is
lost (dead weight loss).
4-2
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
In practice, it is usually impossible to construct the supply curve for an
industry. Instead, the practitioner can use the available cost information,
often for an "ideal" plant, and assume the supply curve is horizontal, in
most cases/ the loss in accuracy caused by making this assumption is not a
severe limitation because the estimated compliance costs, the largest compo-
nent of opportunity - cost, are less affected than the dead weight loss to
society.
Under most circumstances, market prices of goods and services provide a
good estimate of their social opportunity costs. However, in some cases,
there can be significant differences between market price and social cost. As
.noted in Chapter 2, the most significant case relevant to water quality pro-
grams is the difference between social and private rates of return on capital
due to corporate income taxation. In addition, subsidies and grants may
cause the financial costs to participants to differ from social costs.
4.3 MEASURING COSTS: TWO GENERAL APPROACHES
Evaluating the costs of regulatory actions requires the collection and
analysis of relevant industry and municipal cost data. A thorough accounting
of costs for potentially affected firms and cities is desirable. There are two
main approaches to estimate these costs: econometric cost estimation and engi-
neering cost estimation.
Econometric cost estimation, sometimes referred to as the statistical ap-
proach, uses cost-output relationships that are identifiable through empirical
testing. Sometimes, statistical cost estimation offers a way to determine the
costs of proposed alternatives. It is possible that data can be gathered for
actual firms, most of which have identifiably different production processes
and some of which already meet regulatory alternatives of the type under con-
sideration. Given sufficient data, production relationships representative of
complying and noncomplying firms can be estimated. Then, if the prices of
labor and equipment are known, it is possible to establish cost-output relation-
ships known as cost functions. These cost functions, in turn, can be used
to evaluate the cost of regulatory alternatives.
Unfortunately, empirical studies of this type are rarely practicable for
regulatory analyses. The major difficulty usually is that sufficient technical
data are .not available. In particular, published data are usually scarce, and
potentially useful data, available only from firms, are typically considered
proprietary.
The second approach to cost estimation is the engineering cost approach.
This approach offers a viable alternative to statistical cost estimation because
it does not rely on the availability of a firm's actual data. Rather, engineers
familiar with relevant industrial processes use a wide variety of information to
establish relationships between inputs, outputs, and costs. These relation-
ships are presented for hypothetical facilities both with and without proposed
regulatory controls. The practitioner uses the facility data to determine the
costs of the regulatory alternatives. This general approach is the focus of
this chapter.
4-3
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
4.4 TYPES OF COSTS
This section defines the major categories that can be used in assessing
costs. The basic types of cost are capital costs, operating and maintenance
costs, reinvestment, and salvage costs. The discussion is oriented toward
the incremental costs of a regulatory action that affects water quality. In
many situations, this orientation will mean upgrading an existing facility
rather than constructing a new facility.
Incremental costs may include both additional end-of-pipe treatment units
and modification of existing production or treatment units. Most documents
relating to treatment costs are concerned with the former because the set of
unit treatment processes is fairly well defined, allowing cost estimation to be
more standardized. Changes in production processes (or manufacturing) are
more difficult to analyze and are not discussed.
Capital Costs
Capital costs (K) represent initial costs associated with the construction
or upgrading of a facility to meet the treatment requirements plus periodic
reinvestments as individual components wear out and must be replaced.
Table 4-1 shows the kinds of costs that should be included in a capital cost
estimate. They are divided into three main categories: component installed
construction costs, noncomponent construction costs, and nonconstruction
costs. The first category includes physical treatment units (e.g., activated
carbon, chemically assisted clarification) and miscellaneous structures. The
second category includes construction items not necessarily associated with
individual structures, such as site preparation. The last category includes
all the miscellaneous costs in addition to construction costs, including contrac-
tor fees and interest payments. Care must be taken to identify incremental
costs associated with the particular water quality decisions. For example, up-
grades of existing facilities may not require any additional or miscellaneous
structures.
In addition to initial capital costs, replacement costs or reinvestment
costs are required over the life of the project as individual pieces of equip-
ment reach the end of their useful life. Although the line between replace-
ment and repair can be a fuzzy one, the definition is quite operational.
Items that are depreciated over a number of years rather than expensed imme-
diately as costs are considered to be capital items and are treated as reinvest-
ment.
In most cases, no adjustment is required to use .engineering cost esti-
mates as measures of social cost as long as they are based on market prices.
This includes the proper allocation of interest and contractor's fees. Some
caution is required in dealing with reinvestment because the available meas-
ures are based on what tax laws allow and not on the actual social cost.
Volume II will cover this in more detail.
4-4
-------
1
1
Table 4-1 . Components of Capital and Operation and Maintenance Costs
1
1
1
A. Capital Costs
(1) . Component installed construction costs
unit processes
miscellaneous structures
(2) Noncomponent construction costs
site preparation
piping
electrical
instrumentation
(3) Nbnconstruction costs
land costs
engineering and construction supervision
contingencies
.administrative and legal
miscellaneous nonconstruction labor (testing, etc.
design
fees
interest during construction
B. Operation and Maintenance Costs
.
(1) Variable operating costs
• labor
materials
chemical
energy
(2) Byproduct, other credits
(3) Overhead items
insurance
taxes
administrative and other allocations
Operation and Maintenance Costs
:
Operation and maintenance (O&M) costs represent the annual
running and maintaining the facility after its construction (see Table
are divided into three groups: variable operating costs (labor,
energy, etc.)/ byproduct and other credits, and overhead items.
.)
costs of
4-1) and
materials,
Control
efforts resulting from in-plant process changes may also affect revenues or
1
1
1
production costs; these are included in the byproduct credits.
4-5
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
To be consistent with the definition of incremental costs, the overhead
items should include only the additions required by the project, not a pro
rata allocation based on overall costs. The treatment of taxes in a benefit-
cost assessment raises some questions. Although taxes are not costs from the
point of view of the nation as a whole, tax receipts that leave a particular
State may be regarded by that State as a real cost in its benefit-cost assess-
ment. State and local taxes should not be counted as social costs in any
case; they are simply transfers.
Annual O&M costs may vary over the life of the facility. For example,
growth in population or water use will increase the flow to a POTW over time,
causing O&M costs to increase. If the growth rate is high, using the first
year's O&M costs may significantly underestimate average annual O&M costs
over the life of the project.
Reinvestment Costs
Reinvestment costs (RC) represent periodic replacements of individual
units whose lifetime is shorter than that of the overall project. ; Depending on
data availability, it may be more convenient to represent this process as
either the replacement of particular units at discrete intervals or as a con-
stant fraction of initial investment costs each year. Table 4-2 shows the for-
mula for each. If the first approach is used, capital costs may be broken
down into groups with different average ages and the formula applied sepa-
rately to each group using the lifetime applicable to that group.
Salvage Value
Salvage value (SV) is the market value of the facility at the end of the
planning period. A wide range of values is possible depending both on what
that alternative use is and on what assumption is made about reinvestment
(see the previous subsection). One extreme case is that the facility is ex-
pected to continue operating beyond the end of the planning period in the
same fashion as before. In that case, the salvage value depends on the ini-
tial investment cost and the remaining useful life after the planning period.
If the reinvestment process is best characterized as a series of periodic rein-
vestments, the value of the facility is proportional to the ratio of its useful
life at the end of the planning period and its total useful life.
- If the reinvestment process is better represented as an average reinvest-
ment of amount dK each year (see Table 4-2), the expected lifetime of the
facility has no end as long as the annual reinvestments are made. Therefore,
the salvage value at the end of the planning period is still K if the facility is
expected to continue in the same use as before.
If the facility is not expected to continue in its present use after the
planning period, its scrap value must be determined. Any permanent fixtures
4-6
-------
1
1
1
1
1
1
1
•
1
1
1
1
1
•
Table 4-2. Variables and Definitions for Measuring Costs
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
Variable
Planning period
Real social discount rate
Investment costs
O&M costs
Investment lifetime
Physical depreciation rate
Number of replacements in
planning period
Growth rate of O&M costs
Reinvestment costs
Fraction of salvageable value
Salvage value
Present value of stream of
payments for N years at
discount rate s
Present value of rein-
vestment costs
Present value of salvage
Present value of O&M costs
Present value of all
capital costs
Present value of all
project costs
Total annual cost of project
aFirst definition in the right-hand
ments, where RC = K; the second
Symbol
N
s
K
OM
L
d
M
9
RC
q
SV
PV(N,s
PVRC
PVSV
PVOM
PVK
TPV
TAG
Source
Parameters
Parameters
Project costs
Project costs
Project costs
Project costs
Greatest integer N/L
Parameter
K every L years; or
dK every year
Project costs
qK
) (1-(1+s)"N)/s
RC I (1+s)~J ; or
j=1
RC-PV(N,s)
SV(1+s)~N
OM-PV(N,s)
K + PVRC - PVSV
PVK + PVOM
TPV/PV(N,s)
column represents periodic reinvest-
definition
represents annual reinvest-
ments, where RC = dK.
1
1
4-7
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
such as concrete tanks or structures are likely to have ho salvage value.*
Machinery and equipment items may have 25 to 50 percent of their original
value. Cash or natural resources on hand can be valued at 100 percent of
current value, and other current assets may .be valued at 70 to 100 percent
of their current value, depending on their salability.
Summary Cost Measures
Discounting allows the four types of costs to be combined to obtain a
single overall cost estimate for a project. Based on the discussion given in
Chapter 2, the use of the discount rate is summarized below. Specifically/
discount rates are used to construct two kinds of summary measures: pres-
ent value (PV), where O&M costs are capitalized, and total annual costs
(TAG), where capital costs are annualized. Table 4-2 shows these two
methods along with the variables used to develop measures of social cost. In
addition, Table 4-2 summarizes the formulas for the components of the total
present value of the costs of a project.
The present value of all capital and O&M costs (TPV) is obtained by
adding the present values of the cost components (salvage value is sub-
tracted ):
TPV = K + PVRC - PVSV + PVOM
(4.1)
The total annual cost (TAG) of the project is a constant amount whose
present discounted sum over the project period is equal to the present value
of project costs TPV. By the definition of the present value factor (PV)
given in Table 4-2,
TAG = TPV/PV(N,s)
4.5 PRACTICAL ISSUES IN MEASURING COSTS
(4.2)
This section discusses the practical aspects of measuring costs. It
covers sources of data, the use of cost indexes, and major factors affecting
cost estimates. A sample data form, one method of organizing the costing
process, concludes the section.
Sources of Cost Data
There are three commonly used sources of cost data: vendor informa-
tion, estimating manuals, and industry information. In the past, many prac-
*Firms can write off such assets as tax losses. However, these writeoffs
are not appropriate for measuring social costs because society still bears the
full cost of the resources. The distribution of who in society bears the cost
is different. Taxpayers and the firm bear the costs when they are written
off, and the firm and consumers of its product bear the cost when they are
not written off.
4-8
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
titioners have been successful in obtaining cost estimates for both installed
capital costs and annual operating costs from equipment vendors. These esti-
mates are generally solicited in writing along with operating features of the
model plants for which cost estimates are needed.
Estimating manuals have also been useful in developing cost estimates.
Two popular manuals are published by Richardson Engineering Services, Inc.
[1977] -and R. S. Means Company, Inc. [1981]. Richardson's manual is useful
in preparing cost estimates for process industries, and Means' manual is
geared to the construction industry. Both manuals are written to provide
costs for individual components rather than total systems and thus require
some skill and care in their use. Estimates can also be obtained by the use
of cost factors. In this procedure, major equipment costs are multiplied by
appropriate factors to estimate other cost elements. The factors are derived
from experience with previous plant construction costs. Some procedures use
a single factor to estimate total capital investment, but greater accuracy can
be achieved from a method such as Guthrie's [1974], which separates labor
and material costs and applies individual factors to each major process item.
Information supplied by the controlled industry may be useful in estimat-
ing costs. If plants can be identified that are already controlled to the level
under investigation, both total installed costs and annual operating costs can
be obtained from this source. Obtaining estimates of each cost element from
more than one source is a way to validate estimates.
The Use of Cost Indexes
Treatment cost indexes allow cost estimates from different years to be
converted to dollars of a single year to yield a valid comparison. Costs of
various components included in both capital and O&M costs change over time
due both to overall changes in the price level (inflation) and to changes in
relative prices. The problem of forecasting general price inflation can be
avoided by expressing all costs in constant dollars of a given year, but
changes in relative prices still need to be predicted. Unless there are com-
pelling reasons to do otherwise, it is simplest and reasonably safe to assume
constant real costs in future years. Of course, this assumption must be
applied to all components of the benefit-cost assessment to ensure consist-
ency.
Cost indexes are usually represented as a number showing the ratio of
the cost of a unit in dollars of a given year to the cost of the same unit in
dollars of the base year multiplied by 100. In adjusting these costs to con-
stant dollars in a given year, the practitioner must take account of the
change in the index and the overall change in the price level. For example,
if the cost of a treatment plant is given for 1975 in current (i.e., 1975)
dollars and must be converted to 1977 costs in 1977 dollars, the U.S. Environ-
mental Protection Agency's (EPA) Sewage Treatment Plant Construction Cost
(STPCC) Index [Michel] can be used as follows:
4-9
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
.Cost (77) = Cost (75)
= Cost (75) x
STPCC (77)
STPCC (75)
278.3
250.0
= Cost (75) x 1.11 .
If the costs for a given year are to be expressed in the constant dollars of
another year, then the gross national product (GNP) deflator or another
general price index must be used. For example, costs shown above for 1977
may need to be converted to 1976 current dollars. Using the convention that
PGNP (N) is the GNP deflator* in year N and that the notation cost (m, n)
refers to costs corresponding to year m expressed in constant dollars of year
n gives:
Cost (77, 76) = Cost (77, 77)
= Cost (77, 77)
PGNP (76)
PGNP (77)
1.321
1.398
= Cost (77, 77) x 0.945 .
Cost indexes are available for both capital and O&M costs. One index to
use for O&M costs is the U.S. Environmental Protection Agency's (EPA)
Municipal Wastewater Treatment Plant Operation and Maintenance Cost Index
[Michel]. This index is a weighted average of cost indexes for labor, chemi-
cals, power, maintenance, other costs, and a "quality added" factor. Al-
though developed primarily for secondary treatment plants, the mix of inputs
for O&M costs of advanced treatment plants should not differ much.
Several construction cost indexes are available for adjusting capital
costs. These include EPA's STPCC index used earlier, the Engineering
New-Record Cost Index (ENR),t and the Chemical Engineering Plant Con-
struction Cost Index (CE).f The nature of the treatment system being exa-
mined determines which of these indexes is most suitable to use. 'Both the
STPCC and ENR indexes are more oriented to secondary treatment plants,
.where large concrete tanks play an important role, but they are also appro-
*The GNP deflator is published by the U.S. Bureau of Economic Analysis.
The wholesale price index (WPI), another useful index, appears in the U.S.
Bureau of Labor Statistics, Producer^ Pnces and Price Indexes.
tAppears weekly in the Engineering News-Record, published by McGraw
Hill.
^Appears in Chemical Engineering, published by McGraw Hill.
4-10
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
priate for some advanced treatment (AT) processes (such as chemically
assisted clarification), which rely on large tanks. The CE index is more
suitable for processes where equipment plays a large role in costs, such as
use of activated carbon. However, it is difficult to make hard and fast
rules. The cumulative percent increases of the STPCC and ENR indexes are
close, while that of the GE index is somewhat lower.
To arrive at a cost figure more accurate than one that results from the
use of the national average indexes alone, locality factors can be applied to
an estimated cost or cost index. The use of locality factors, which have been
calculated from generally available statistics, permits the localizing of national
average cost data on various cost items. Locality factors are available from
the U.S. Environmental Protection Agency's (EPA) Sewage Treatment Plant
and Sewer Construction Cost Index [Michel].
Major Factors Affecting Cost Estimates
The cost of a treatment process is influenced by a multitude of factors.
Among the important ones are wastewater flow rate, pollutant loadings, plant
location, and performance (i.e., amount of pollutant removed and effluent
concentrations). Most treatment technologies show economies of scale; i.e.,
costs increase at a slower rate than flow size. Thus, the cost per gallon of
wastewater treated in a large plant is less than that treated in a smaller plant
that meets the same treatment performance criteria. In general, the scaling
factor* varies with size of flow and technology. For preliminary planning-
purposes, the following values are- useful for extrapolating the treatment cost
of a given treatment plant size to others:
Cost Item Scaling Factor
Capital 0.6 to 0.9
O&M 0.7 to 0.9
Labor 0.5 to 0.7
Utilities and 1.0
chemicals
The upper end of the range for the scaling factors (implying less economy of
scale than the lower end) is associated with the more advanced treatment tech-
nologies, such as carbon adsorption and electrolysis, to which this handbook
is oriented. Economies of scale- also appear in the relationship of cost to
waste loading but are less prevalent than in the cost-to-flow-size relationship.
*As used here, scaling factor refers to x In the following equation:
where Cj is a cost for a treatment plant with flow Qj and C2 is the estimated
cost for a plant with flow Q2.
4-77
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Because wastewater flows and loadings tend to show greater variability
in a POTW than they do in an industrial plant, equalization is commonly used
to smooth out fluctuations. Even so, the POTW customarily is designed with
some excess capacity to accommodate variations in flow and waste loading.
In some cases, more intensive treatment effort in the treatment plant—
such as additional chemicals or energy input—can result in enhanced perform-
ance or accommodation of a greater volume of wastewater without sacrificing
design performance. However, the incremental cost usually is high. There-
fore, .it is less costly in the long run to upgrade or expand the treatment
facility if actual flow, loading, or performance is expected to change appre-
ciably from initial design conditions.
It is generally more costly to upgrade an existing plant by retrofitting
than it is to use the same treatment train in a new plant. The treatment in
place at a plant may limit the choice of higher treatment technologies that can
be selected to upgrade the plant. Upgrading usually requires engineering
effort, rewiring, and additional piping at the existing facility. The additional
cost attributed to retrofitting a plant is sensitive to the specific features of
the plant and the site, so these factors should be considered in developing
retrofit .cost factors. Based on limited observations, costs of retrofitting an
existing plant can range from 1 .to 15 percent higher than the cost of incor-
porating the same treatment train in the initial design of a new plant [U.S.
EPA, 1976]. The added cost of retrofit in percentage terms is inversely pro-
portional to the capacity of the treatment facility.
Sample Data Form
A "model plant form" can be used to compile parametric and cost esti-
mates in an orderly fashion. The form can be continually revised to meet the
needs of the specific data requirements of an assessment. A model plant
questionnaire of this type is applicable only for regulatory alternatives that
involve engineering controls. Figure 4-2 presents a version of the model
plant questionnaire.
4.6 COST ESTIMATING TECHNIQUES
This section on cost estimating techniques defines the components of
treatment systems and describes several variations on the basic costing tech-
niques. Finally, the question of how to proceed when no directly relevant
cost sources are available is discussed. The techniques described here are
appropriate for costing both POTWs and industrial dischargers.
Components of Treatment Systems
To analyze the costs of a particular proposed treatment system, it is
useful to break it down into its components. Although each treatment system
has its unique aspects, the individual components are more standardized and
hence more easily costed using standard references, where costs are defined
in terms of a few major parameters. The detail of the breakdown depends on
the accuracy of the estimate required. It is useful to understand the relation-
ships of the following three levels of treatment units:
4-12
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
MODEL PLANT FORM
1. Model Ham Number and Description
Number
Description
2. If new, ran design tif* (yeari)
If existing, itata plant age (yean)
and remaining life (yean)
3. Legal Depreciation Period (yean) and Usual Depreciation Method
Plant
Compliance Equipment
4. Discuss what relationship exirti, if any, between existing plant age and model type. Discuss age distribution of exist-
ing plants.
Alternative 1
(Baseline)
Alternative 2
Alternative n
5. Product Names, Annual Capacity Outputs
Primary Product
Name
Annual Physical Output Capacity
Units Specified
Producer Price, 19_
per unit
.dollars
Secondary Product
Name
Annual Physical Output Capacity
Units Specif ied
Producer Price, 19 dollars
per unit
Marketable Byproduct
Name
Annual Physical Output Capacity
Units Specified
8. Annual In-Plant Consumption of any of the
Products Lilted in (5L
Name
Quantity
Name '
Quantity
7. Installed Capital Cost 19 dollars
OF NEW PLANT
8. Installed Capital Cost of Compliance
Equipment for Existing Plant
19 dollars
XXX
Figure 4-2. Sample model plant data form.
4-13
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Model Plant Form (con.)
Alternative 1
(Baseline)
Alternative 2
Alternative n
9, Expected Life of Compliance
Equipment for Existing Plant (years)
10. Lott Productive Time While Compliance
Equipment It Installed in Existing Plant (days)
11. Salvage Value, if any, of Compliance
Equipment When Plant Is Closed
19 dollars
12. Total Annual Operating Costs
19 dollars
13. Annual Fixed Operating Costs
Total Fixed Operating Cost
19 dollars
Labor
Number of labor hours
Annual Wage per labor hour
18 dollars
Labor Cost 19 dollars
Energy Costs 19 dollars
Materials Costs 19 dollars
Overhead Costs 19 dollars
Other Fixed Costs 19 dollars
14. Annual Variable Operating Costs
Total Variable Operating Costs
19 dollars
Labor
Number of labor hours
Annual Wage per labor hour
19 dollars
Labor Cost 19 dollars
Energy Costs 19 dollars
Materials Costs 19 dollars
Overhead Costs 19 dollars
Other Variable Costs 19 dollars
15. Current Salvage Value of Plant
19 dollars
XXX
XXX
XXX
16. Salvage Value of Plant at End of Life
19__ dollars
17. Residuals Discharged to Environment
To Atmosphere
Name '.
Quantity
To Water
Name _
Quantity
To Land
Name
Quantity
Figure 4-2. (con.)
4-14
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Unit process: A unit process corresponds to a single treat-
ment operation. Examples of unit processes are secondary
clarification, filtration, and cyanide destruction.
Treatment process: A treatment process is a sequence of one
or more unit processes linked together to support a particular
pollutant-removal process. For example, the activated sludge
process involves the decomposition of organic pollutants by
microorganisms. This operation requires a number of unit proc-
esses including aeration, sedimentation, and sludge reactiva-
tion.
• . Treatment train: A treatment train is a sequence of treatment
processes. For example, an advanced treatment train may con-
sist of the following treatment processes: preliminary screen-
ing, primary settling treatment, secondary biological treatment,
and nutrient removal by chemical addition.
As discussed below, cost estimates for benefit-cost assessments of water qual-
ity programs should be broken down at least to the treatment process level
and, in some cases, to the unit process level.
Estimating Treatment Costs
Cost estimation requires a specification .of the treatment train . to be
used. A waste treatment train can be described by a flow diagram showing
the relation and function of the various treatment and unit processes. One
way to proceed is to (1) specify the .important design parameters (such as
flow, influent and effluent concentrations, and contact time) of each unit in
the treatment process, (2) calculate the resource requirements (for example,
for site preparation and construction and for purchased equipment, energy,
and labor) of each unit, (3) estimate the indirect costs, and (4) sum to ob-
tain a total cost. This calculation is carried out separately for a total capital
and a total O&M cost before they are combined into a present value or annual -
ized cost.
Before illustrating how different sources of cost information can be
applied in particular situations, it is useful to distinguish among four differ-
ent approaches that are employed to make cost estimates:
A total system estimate
A planning level estimate
An engineering estimate
A contractor estimate.
The cost estimates produced by these techniques range from gross to refined,
depending on the different stages of a project, which range from project con-
ceptualization to request for contractor bids. The four costing techniques
4-15 U.S. EPA Headquarters Library
Mail code 3201
1200 Pennsylvania Avenue NW
Washington DC 20460
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
provide a convenient frame of reference for discussion purposes .but do not
imply that rigid distinctions can always be made in using one technique
versus another in a specific situation.
The use of site-specific information should result in more accurate
results than use of generalized information. While it is, of course, desirable
to develop accurate estimates, the practitioner is always confronted with decid-
ing how much accuracy is needed for a particular phase of project planning
and what level of effort to commit to the development of the estimates. For
the purposes of water quality standards planning, the planning level estimate
is generally the most appropriate one to use, and the examples focus on that
level.
Total System Estimate. In contrast to the other three costing
techniques, the total system approach does not attempt to par-
tition a treatment train into treatment or unit processes.
Usually only one parameter of the treatment system—for
example, plant capacity expressed as daily flow—is used with a
set of cost curves to obtain total capital and annual O&M
costs. The accuracy is ±40 percent.
Planning Level Estimate. This approach is based on prior
analyses of treatment system components or unit processes in
which costs of the units have been related to important design
parameters. The purpose is to allow recombination or syn-
thesis of total costs resulting from any combination of the unit
processes using specified values ,of the design parameters.
This level of estimate is appropriate for most water quality pro-
gram planning purposes. Application of this technique re-
quires that the practitioner identify the major components in
the wastewater treatment train, the associated design param-
eter values, and the access to generalized cost functions for
the components. The accuracy of this costing technique is
within ±30 percent. The practitioner may be able to improve
the accuracy of the estimate if judgments can be made about
how site-specific characteristics differ from average conditions
embodied in the generalized cost functions,
Engineering Estimate. Like the planning level estimate, the
engineering estimate is calculated using unit process data but
goes into more detail on the unit processes in the system to
adjust specific costs. This technique should yield a cost esti-
mate within ±15 percent.
Contractor Estimate. The contractor estimate is based on
specific engineering designs—or design approaches coupled
with specified performance requirements--for the treatment
system and it's unit processes. The precision of the contrac-
torfs estimate should be within ±5 percent.
4-76
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
4.7 EXAMPLES USING COST TECHNIQUES
Detailed engineering cost estimates are unlikely to be available for analy-
ses of the type described in this handbook. Therefore, this section explores
a range of techniques, from fairly sophisticated costing models (CAPDET) to
process handbooks and EPA development documents, and presents specific
examples to show how they are applied. Their applicability depends on both
the amount of information and the t'ime and resources available to the analyst.
States may have their own techniques that are equally applicable. Even if
specific information is not available, there are basic similarities in the ap-
proaches to water treatment problems taken by different industries. There-
fore,' the experience of other industries with similar processes or pollutants
may provide a reasonable guide for estimating costs.
A library of cost information* is available to the practitioner, including
documents and computer programs. The purpose of the following examples is
to illustrate the use of various information sources to estimate costs, particu-
larly planning level estimates.
JExampjeijiii1_:__rfiPijlp__M}|.| Using EPA peyelojpmejTt^Dqcument
This example illustrates the use of an EPA Development Document [U.S.
EPA, 1980b] for a specific industry and the planning level costing approach.
Assume that more stringent effluent standards will be imposed on a 900-ton-
per-day kraft pulp mill to meet water quality standards. Based on a waste-
load allocation, the effluent concentrations required to meet the water quality
standards are 15 mg/L for both biochemical oxygen demand (BOD) and total
suspended solids (TSS). The costs are to be estimated for October 1981.t
The practitioner determines that the mill belongs to the "Market Bleached
Kraft" subcategory (described as one of the industry subcategories in the
Development Document) and that concentrations of BOD and TSS in the
plant's effluent stream currently are 20 mg/L and 30 mg/L, respectively,
which meet Best Practicable Technology (BPT) standards.^ The Development
Document identifies three treatment options that might possibly be used to
meet the new BOD and TSS targets, but only Option 3, which consists of
additional in-plant process controls together with chemically assisted clarifica-
tion of the final effluent, can achieve the BOD and TSS targets of the water
quality standards.
*lmportant sources include: (1) EPA Development Documents for effluent
limitations guidelines and standards (issued by the Effluent Guidelines Divi-
sion of EPA to provide the technical background for the development of waste
treatment rules for particular industries); (2) Areawide Assessment Proced-
ures Manual [U.S. EPA, 1976]; (3) Innovative and Alternative Assessment
Manual [U.S. EPA, 1980a].
fThe examples in this guidance use historical values of treatment cost
indexes. For planning purposes it may be necessary to estimate future
values.
standards for the pulp, paper and paperboard industry have not
yet been promulgated.
4-17
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
The next step is to estimate the cost for Option 3. Table 4-3 shows the
costs of meeting Option 3 for three model mills with sizes of 350, 600, and
1,600 tons per day. Note that energy costs are presented separately, so
they must be combined with O&M costs to obtain the definition of O&M costs
used here. Two costs are shown for each category of cost based on alum
concentration. As a first approximation, an average cost for the two alum
concentrations is used. Next, because none of the three model mill sizes is
Table 4-3. Example 1: Cost Summary—
Market Bleached Kraft, Subcategory
Mill size
(tons/day)
Capital cost 350
600
1,600
Annual O&M cost 350
600
1,600
Annual energy cost 350
600
1,600
Incremental
compliance costs .
from BPT: Option 3 '
6,662
8,974
9,446 ,
16,590
17,410
947
1,327
1,327
1,953
2,974
4,550
212
517
351
358
897
917
SOURCE: U.S. EPA [1980b], p. 468, Table IX-7.
aFirst quarter 1978 thousands of dollars.
Dollar value shown above the line is based on chemical assisted clarifica-
tion dosage of alum at 150 mg/L; the value below the line is for dosage at
300 mg/L.
4-18
-------
I
I
mm .
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
for a 900-ton-per-day plant, a linear interpolation based on mill size is made
between the values for the 600- and 1,600-ton-per-day plants.* This yields a
capital cost for Option 3 of $11.5 million. Annual O&M cost is $2.3 million,
and energy cost is $0.5 million; these are added for a total O&M cost of
$2.8 million.
The following steps show the results for various measures of social cost
derived from the basic costs, which are used as part of the benefit-cost
.assessment. The following parameter values are used:
ENR index (1st Qtr 1978) =2,683
ENR index (Oct 1981) = 3,679
EPA O&M index (1st Qtr 1978) = 2.30
EPA O&M index (Oct 1981) = 3.34
s = real social discount rate = 0.10 or O.OSf
L = average equipment lifetime = 15
d = depreciation rate = 1/L = 1/15
N - planning .period = 30 years
The indexes are historical values, the other parameters are assumed values.
The results are (all quantities in thousands of 1981 dollars unless otherwise
noted):
1. Capital treatment costs:
K = 11,500 (1st Qtr 1978 $)
K = 11,500 xfigf
K = 15,800 (Oct 1981 $)
2. O&M treatment costs:
OM = 2,800 (1st Qtr 1978 $)
*Another method for estimating the value for the 900-ton-per-day plant
size is to fit a cost curve to the three data points to determine if there are
economies of scale of treatment costs instead of assuming a linear relationship
between mill size and cost for the two larger model mills.
TThese are just sample values chosen to show the effects of different
values on the results. See Chapter 2 for discussion of which discount rate to
use.
4-19
-------
1
1
^^B
I
•
1
1
1
1
•
•
1
1
1
1
1
1
1
1
3 34
OM = 2,800 x |^|2
2.30
OM = 4,100 (Oct 1981 $)
3. Present value factor (equals present value of payment of one dollar
per year for N years at discount rate s):
PV(30, 0.10) = 9.43
PV(30, 0.05) = 15.37
4. Present value of O&M costs:
PVOM = OM • PV(30, 0.10) = 4,100 • 9.43
- OM • PV(30, 0.05) = 4,100 • 15.37
= 38,700
= 63,000
5. Reinvestment cost (assume the entire facility is .replaced in 15 years
at the same original real cost):*
RC = K = 15,800
6. The present value of the reinvestment cost is found by discounting
at rate s over 15 years:
PVRC = RC • (1+s)"L = 15,800 • <1.1)"15 =
= 15,800 • (1.05)"15
7. Salvage value (assume zero salvage value)
SV = 0
8, Present value of salvage (salvage occurs
N):
PVSV = SV (1+s)~N
= 0
9. Present value of salvage (salvage occurs
N):
*
*Not all structures would need to be replaced
to represent the average lifetime of the facility.
4-20
= 3.800 (s = 0.10) or
= 7,600 (s = 0.05)
*
at end of planning period,
at end of planning period,
. Fifteen years is assumed
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
PVSV = SV (1 + s)"N
= 0
10. Total present value of investment cost:
PVK - K + PVRC - PVSV
= 15,800 + 3,800-0 = 19,600(5 =
= 15,800 + 7,600 - 0 = 23,400 (s = 0.05)
11. Total present value of project cost:
TPV = PVK + PVOM
= 19,600 + 38,700 = 58,300 (s = 0.10)
= 23,400 + 63,300 = 86,400(s = 0.05)
12. Annualized capital costs:
KANN = PVK/PV = 19,600/9.43 « 2,080 (s = 0.10)
= 23,400/15.37 = 1,520(5 = 0.05)
13. Total annualized costs:
TAG = KANN + OM = 2,080 + 4,100 = 6,180 (s = 0.10)
= 1,520 + 4,100 = 5,620 (s = 0.05)
The total present value of project costs calculated in Item 11 above is the
amount that would be included in the benefit-cost assessment. Alternatively,
the total annualized cost (Item 13 above) would be used if the rest of the
assessment were also expressed in terms of annualized costs.
Example 2: POTW Using CAPDET
Over the past decade a number of computer-based treatment cost esti-
mating models have been developed; CAPDET is one of the more widely used
models [U.S. EPA, 1981].* CAPDET is not a mathematical optimization model;
the CAPDET approach is to prepare cost estimates for alternative treatment
trains specified by the user. With reference to the four costing techniques
discussed earlier, the CAPDET method is probably best described as an inter-
mediate method between the planning level approach and the engineering esti-
*To obtain access to CAPDET programs and documentation, contact the
Systems Analysis Group of EPA regional offices or the Facilities and Require-
ments Division, Office of Water Program Operations, EPA in Washington, D.C.
4-21
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
mate approach. CAPDET contains cost and performance equations for-35 unit
processes such as activated sludge, carbon absorption, lagoons, incineration,
etc. For some unit processes, alternative design approaches are included;
for example, the activated sludge process is described by 13 different meth-
ods of aeration. Two separate cost estimating methods are incorporated in
CAPDET. First, parametric cost estimating is based on statistical analysis of
the cost of facilities of similar size and characteristics at other locations.
Second, unit cost estimating is based on identification of cost elements to
which input prices are applied—e.g., cubic yards of concrete in a clarifier
are quantified and an input cost value for reinforced concrete is applied to
obtain a construction cost.
After recent revisions, CAPDET can be used to estimate costs of upgrad-
ing an existing POTW even though the program was developed originally to
estimate new plant costs. Its capabilities have also been expanded to provide
estimates of financial impacts on households from the construction of a POTW
facility.
In this example, an existing municipal POTW has an average daily flow of
5 Mgal/d, a maximum flow of 10 Mgal/d, and effluent concentrations of 15 and
20 mg/L for BOD and TSS, respectively. Current secondary treatment uses
plug flow for the activated sludge effluent, and the practitioner needs to
determine the cost of upgrading the treatment by chemical addition and filtra-
tion to meet the water quality standard that requires reductions of BOD to
10 mg/L and TSS to 1 mg/L or less.
To use the CAPDET computer program, the unit processes in the current
plug flow system and the upgraded treatment are specified as shown in Table
4-4. The CAPDET program can accept any number of wastewater treatment
trains that are suitably described by the user.
Table 4-4, Example 2: Current and Upgraded
Treatment Trains for CAPDET
Current treatment sequence
Upgraded treatment sequence
Raw sewage
Preliminary treatment
Primary clarifier
Plug flow
Secondary clarifier
Chlorination
Gravity thickener
Anaerobic digestion
Vacuum filtration
Hauling and land fill
Raw sewage
Preliminary treatment
Primary clarifier
Plug flow
Secondary clarifier
Coagulation
Filtration
Chlorination
Gravity thickener
Anaerobic digestion
Vacuum filtration
Hauling and land fill
4-22
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Table 4-5. Example 2: Price and Cost Inputs to CAPDET
Cost analysis input parameters
Interest rate: 10.000 percent
Planning period: 30 years
Year of dollars used: 1980
Cost Indexes
Buildings
Excavation
Wall concrete
Slab concrete
Marshall and Swift Index
Crane rental
EPA Construction Cost Index
Canopy roof
Labor rate
Operator class 11
Electricity
Chemical costs
Lime
Alum
Iron salts
Polymer
Engineering News Record Cost Index
Handrail
Pipe Cost Index
Pipe installation labor rate
8-in. pipe
8-in. pipe bend
8-in. pipe tee
8-in. pipe valve
Unit prices
55.00 $/ft2
7.00 $/yd3
207.00 $/yd3
91.00 $/yd3
577.00
67.00 $/h
163.00
15.75 $/ft2
13.40 $/h
9,00 $/h
0.04 $/kWh
0.03 $/lb
0.04 $/lb
0.06 $/lb
1.62 $/lb
2,886.00
25.20 $/ft
295.20
14.70 $/h
9.08 $/ft
86.82 $/unit
128.49 $/unit
1,346.16 $/unit
4-23
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Table 4-5 shows unit costs and prices in 1980 dollars that the practi-
tioner provides to CAPDET, as well as the planning period and interest rate
to be used in computing project costs. In calculating the capital costs,
CAPDET recognizes that each equipment item has a service life associated with
it that may be less than the planning period and that the facility has some
salvage value at the end of the planning period. As a result, the capita!
cost incorporates three components, initial and reinvestment costs and salvage
value. Typical useful life periods are as follows: wastewater conveyance
structures, 50 years; other structures, 30 to 50 years; process equipment, 15
to 20 years; and auxiliary equipment, 10 to 15 years. In addition, O&M costs
can be specified to vary over the planning period.
Table 4-6 summarizes the costs of the two systems analyzed by CAPDET,
a new treatment plant using the current treatment sequence and a new plant
with the two new treatment steps included. The table also shows the unad-
justed cost increments obtained from CAPDET for upgrading the POTW and
the adjusted costs which are increased by 15 percent to reflect the costs of
retrofitting an existing system.
The capital costs of upgrading (as shown in Table 4-6) are $1.01 million;
total project costs are $1.81 million. Additional O&M costs (which vary be-
tween the first and final year) range from $370,000 to $200,000 annually.
Present worth is.$5.46 million; annualized cost is $580,000 based on a 30-year
planning period and 10 percent rate of interest.
Table 4-6. Example 2: POTW Upgrading Cost Summary
(millions of 1980 dollars)
Incremental
New New POTW with
POTW with plug flow costs for upgraded POTW
plug flow and chemical **
secondary addition and CAPDET Adjusted for
treatment filtration unadjusted retrofit
Capital cost
Total construction
3.60
4.40
4,48
5.46
0.88
1.06
1.01
1.22
cost
Total project
cost
O&M cost
6.53
8.10
1.57
1.81
First year
Final year
Present worth
0.31
0.46
9.98
0.63
0.63
14.76
0.32
0.17
4.78
0.37
0,20
5.46
Adjusted costs are 15 percent higher than unadjusted costs.
4-24
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Example 3: POTW Using Technology Assessment Manual
In this example, planning level estimates of costs for upgrading an exist-
ing facility are developed from generalized cost curves that have been devel-
oped
[U.S.
in the 'EPA's Innovative and Alternative Technology Assessment Manual
EPA, 1980a]. The manual has been designed specifically to aid Federal
and State review authorities in the administration of innovative and alternative
requirements of the Construction Grants Program and to provide basic method-
ological and technical information to individuals involved in facility plan de-
velopment.
A municipality is assumed to have secondary wastewater treatment in
place at the POTW using the activated sludge process. Wastewater flow is 10
Mgal/d, and the effluent has a monthly average value of 30 mg/L for both
BOD 'and TSS. To meet an ultimate oxygen demand (UOD)* of 85 mg/L as
part of waste-load allocation, it has been determined that the advanced waste
treatment process described in the Manual as "nitrification, separate stage,
with clarifier" is required.. The construction and O&M costs are obtained
from Fact Sheet 2.1.14 in the Manual using a 10-Mgal/d flow .rate. Construc-
tion costs are $1,6 million and O&M costs are $70,000 in 1976 dollars.
Several other adjustments to the values read from the cost curves are
necessary. The referenced Fact Sheet is based on an ENR construction cost
index of 2,475 (for September 1976). Adjusting the construction costs for a
first quarter 1980 ENR index of 2,886 yields a construction cost of $1.9 mil-
lion. The O&M costs must also be adjusted to first quarter 1980. The EPA
O&M index for the third quarter of 1976 is 2.06 and for the first quarter of
1980 is 2.83, yielding an adjustment factor of 1.37. Therefore, O&M costs
are $100,000.
The Fact Sheet in the Manual directs the user to estimate other capital
expenditures that have not been included in the construction cost curves.
Table 4-7 replicates Table A-2, which is provided for that purpose in the
Manual. Following the directions incorporated with Table A-2, the practi-
tioner estimates nonconstruction capital expenditures. As shown in Table
4-7, total capital cost is $3.5 million.
The costs of the system are summarized in Table 4-8 together with efflu-
ent information. The annualized costs in the table are based on a capital
recovery factor of 0.106, reflecting a 10 percent social discount rate and a
30-year plant life.
Example 4: What To Do When There Is "No Information"
The practitioner may find that no Development Document or other techni-
cal data have been published that are directly applicable to a particular type
of industrial plant. Nevertheless, it is possible to identify treatment options
and develop preliminary cost estimates, suitable for at least the early phases
of water quality program analysis.
*UOD + BOD5 x 1.5 + NH3 x 4.5.
4-25
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Table 4-7. Example 3: Development of Capital Costs (Nitrification)
(All costs in millions of 1st quarter 1980 dollars)
Component installed construction costs
Unit processes
Nitrification
Miscellaneous structures
(Administrative offices, laboratories,
shops and garage facilities)
Subtotal 1
$1.9
$0.0
$1.9
Noncomponent costs
Piping
Electrical
Instrumentation
Site preparation
Subtotal 2
Nonconstruction costs
Engineering and construction
supervision @ 15 percent.
Contingencies @ 15 percent
Subtotal 3
Total capital costs
Average
10%
•8%
5%
5%
Range
8-15%
5-12%
3-10%
1-10%
$0.2
0.4
0.1
0.1
$0.4
$0.4
$0.8
$0.8
$3.5
Range due to level of complexity, degree of instrumentation, subsoil con-
dition, configuration of site, etc., percentage of subtotal 1.
^Percentage of subtotal 1 plus 2.
Table 4-8. Example 3: Summary of Costs
for Nitrification Upgrade
Costs (millions of 1980:1 dollars)
Construe- Total
tion capital O&M Annual ized
costs . costs costs • cost
($) ($) ($) ($)
Effluent (mg/L)
BOD
UOD*
1.9
3.5
0.10
0.46
10
19.5
UOD = BOD5 x 1.5 + NH3 x 4.5.
4-26
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
A reasonable approach is to first identify the pollutants that are of major
concern at the particular plant, their concentration, and the effluent flow.
The next step is to identify waste treatment unit processes that may be appli-
cable for the identified pollutants. This can be done by identifying one or
more analogous industries--e.g., an industrial plant where the specific pollut-
ants of interest are being effectively treated, or have been studied. In all
likelihood, the method of treatment of a specific pollutant in one industry will
be applicable to the particular plant of interest to the practitioner.
The next step is to acquire Development Documents and other reports on
the analogous industries. Using the information on the analogous plants, the
practitioner should search out the cost-versus-flow-size relationships for the
one or more unit processes used to treat the pollutants of interest. This will
allow the practitioner to synthesize a wastewater treatment train and estimate
a total cost for the process units in the train. This approach is preferable
to using a total system cost estimate, in which the costs for each unit process
are not explicitly identified.
4.8 FINAL CAUTIONS
This section describes several important factors that suggest caution is
necessary when measuring waste treatment costs, particularly industrial
wastes.
The costs of treating wastes are only one element of the entire produc-
tion process for a firm. This element is the management of .the residuals that
accompany production (see Ayres and Kneese [1969]). The costs of treating
wastes can be affected by changes in the level of production, which alters
the volume of wastes and perhaps the type of treatment required. Changes
in the type of product produced, in the processes used to make it, or in the
means of recovering the waste all can affect the cost and type of treatment
required. For example, a technology change in pulp production for the paper
industry from sulfite to kraft lowered the volume of suspended solids. When
the practitioner considers the cost of a regulatory action, it is important to
keep this total system view in mind,
4.9 SUMMARY
This section summarizes the major points from the chapter.
Costs are measured on the basis of opportunity cost—the value
of the next best forgone alternative.
The two main approaches to estimating costs are the economet-
ric and engineering approach. The engineering approach is
most often used because data needed in the econometric ap-
proach are seldom available.
Two major cost categories are capital costs and operation and
maintenance (O&M) costs. Capital costs are initial costs of
4-27
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
construction or upgrading. Operation and maintenance costs
are annual costs of running and maintaining a facility after its
construction.
Three useful sources of cost data are vendors, estimating
manuals, and industry information.
Discounting links the four major components of project costs:
initial investment, Q&M, reinvestment, and salvage value.
Cost indexes allow cost estimates from different years to be
compared on a common basis. Useful index are EPA's Sewage
Treatment Plant Construction Index, the GNP price deflator
series, and Chemical Engineering Plant Construction Cost
Index.
Important factors affecting costs of waste treatment are flow
rate, pollutant loadings, plant location, and performance cap-
ability.
A model plant data form .can organize information .required for
costing.
Three levels of waste treatment are unit process (single treat-
ment), treatment process (sequence of unit processes), and
treatment train (sequence of treatment processes).
Planning level estimates of costs based on prior analyses of
treatment cost and having an accuracy of ±30 percent are ap-
propriate for many water quality program decisions.
EPA Development Documents provide a valuable .source of
planning level costs for specific industries.
CAPDET is a computerized model that prepares cost estimates
of alternative treatment trains, estimates the costs of up-
grading, and computes a financial impact statement for publicly
owned treatment works. CAPDET provides accuracy between
±15 and ±30 percent.
Technology assessment manuals provide basic technologies for
publicly owned treatment works.
Treatment cost estimates are sensitive to many factors in firms'
overall production operation, including output levels, types of
products, or manufacturing processes.
4-28
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
CHAPTER 5
COMPLETING THE BENEFIT-COST ASSESSMENT
5.1 INTRODUCTION
What is a sensitivity analysis? How is it used in a benefit-cost assess-
ment? What methods are appropriate for displaying the results of a benefit-
cost assessment? Is a checklist of elements possible in a benefit-cost assess-
ment?
In answering these questions, this chapter discusses three sets of prac-
tical problems encountered in benefit-cost assessments: establishing plausible
results, displaying those results, and organizing the assessment's elements.
The complexity of these practical problems varies directly with the complexity
of the particular assessment. For example, in a qualitative assessment, a
sensitivity analysis need be conducted only in very general terms. However,
in a complex assessment, such as one presented in the example below, the
plausibility of individual variables is specifically considered. In other words,
the resources used in a sensitivity analysis can be tailored to the importance
of the decision.
The following sections of this chapter highlight these practical aspects of
a benefit-cost assessment. Specifically, Section 5.2 presents a sensitivity
analysis for a water quality standards example involving monetized benefits
and costs. Section 5,3 describes narratives, arrays/ and graphs as alterna-
tive ways of displaying the results of an assessment. Section 5.4 presents a
checklist for organizing an assessment. Finally, Section 5.5 summarizes the
chapter's main points.
5.2 SENSITIVITY ANALYSIS: A GAUGE TO BELIEVABILITY
Introduction
One ingredient in a good benefit-cost assessment is a sensitivity analysis
of its key variables and assumptions. The most common variables considered
in a sensitivity analysis are the parameters that determine benefits and costs,
the discount rate, and the time horizon of the assessment. For example, the
effectiveness of a particular treatment process may be uncertain so that the
anticipated water quality may not be fully achieved. As previously discussed,
the discount rate can be among the most important of these features because
it affects both benefits and costs.
A sensitivity analysis establishes a range for the net benefits in the
assessment rather than simply portraying a single estimate. In principle, this
is similar to the procedure in statistics that establishes interval estimates to
5-1
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
bound the range of possibilities. In a benefit-cost assessment, greater uncer-
tainties in the estimates of benefits and costs yield larger bounds to include
most of the possible outcomes.
A sensitivity analysis will show when the assessment is affected by the
assumptions made. In instances where similar benefit-cost assessments are
expected in the future, a sensitivity analysis can serve as an agenda for re-
search by highlighting assumptions that influence the estimated net benefits.
Even when an assessment is sensitive to the assumptions employed, its results
are not invalidated. Rather, this sensitivity calls for more care in interpret-
ing the results and in determining whether the assumptions are reasonable.
A sensitivity analysis employs high and low estimates for both benefits
and costs and estimates net benefits for a range of discount rates. Although
there are no formal procedures in a sensitivity analysis, the following example
highlights the decisions required to implement the various steps.
Table 5-1. Key Elements of Benefit and Costs
Line item
(and key
assumption)
Cost savings3
Likely estimate
(and range),
$ million
Benefits forgone
Line item Likely estimate
(and key (and range),
assumption) $ million
Capital costs
New advanced waste
treatment plant
(size of plant) 4,8 (3.0 to 6.0)
Process changes
at meat processing
plant (extent of
changes)
3.2 (1.2 to 5.0)
Additional fish-
ing (5 percent
growth rate)
Additional swim-
ming (probability
of swimming is
constant)
1.75 (0.50 to 3.0)
0.50 (0.2 to 1.25)
Additional near-
water activities
(no new activities) 0.50 (0.2 to 1.50)
Operating costs
Advanced treatment
Meat processing
1.0
0.5
Cost savings are the investment and operating costs forgone by not meeting
the fish and wildlife propagation use for river. Likely estimate is listed first
with the range in parentheses.
Benefits forgone stem from the recreational activities had the fish and wildlife
propagation use been achieved. Likely estimate is listed first with the range
in parentheses.
5-2
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Example
Suppose a State is considering changing the use designation for Segment
30 of a river from fish and wildlife propagation (which is not being attained)
to agricultural uses. The need for a sensitivity analysis is demonstrated by
the data in Table 5-1, which shows the key elements in the benefit-cost as-
sessment--cost savings and benefits forgone. Uncertainty both in the ulti-
mate size of a new municipal waste treatment plant and in the estimation of
the cost savings for the process change at a meat packing plant dictates the
bounds for the cost estimates. In this case, the capital cost savings occur in
the current year, so they are not affected by the selection of the discount
rate.
The estimates of the benefits forgone by the change in designated uses
show even more uncertainty. The value of the estimated loss in fishing activ-
ities depends on a 5 percent per year increase in fishing expected under the
previous designated uses. By varying the assumed increase in fishing activi-
ties under the new designated uses, the forgone benefits range from $500,000
per year to $3,000,000 per year. The estimation of forgone swimming activ-
ities is based on the assumption that the level of swimming would not change.
If adjustments are made for the uncertainty of the assumption, the estimated
forgone swimming benefits range from $200,000 per year to $1,250,000 per
year. The range of forgone benefits for near-water activities depends on the
assumption that no new activities are developed for the river. When this
assumption is relaxed, benefits for near-water activities range from $200,000
to $1,500,000.
Step 1: Translate the Benefits and Costs into Present Values
The first step in sensitivity analysis is to translate the benefits and
costs into present values to make the net benefit calculation. This example
simplifies the calculations by assuming a real rate of 4 percent and a project
-life of 50 years. Only quantified and monetized benefits are considered in
this example, and all capital costs are spent in year zero. The problem;
then, is to translate the annual ,costs and annual benefits that occur each
year into their present value equivalents, which requires the use of present
value tables. The information needed for using present value (PV) tables is
P/A,* 4 percent, 50 years. For the most likely case:
Operating cost savings = P = $1.5 (21.482) = $32.2 million
Total cost sayings = 32.2 million plus 8.0 million capital cost for-
gone, or $40.2 million
Benefits forgone = P = $2.75 (21.482) = $59.1 million
*P/A is a heading found in most present value tables for translating
annual costs into their present value equivalents.
5-3
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Net benefits calculated for the most likely estimate benefits and costs are:
Net benefits = PV benefits forgone - PV cost savings
$18.9 million = $59.1 million - $40.2 million
The assessment of net present value benefits shows the cost savings from the
change would be outweighed by the benefits forgone if the most likely esti-
mates are used. Table 5-2 summarizes -the results of calculations for the
ranges of benefits and costs in addition to the most likely case.
*
Step 2: Perform Sensitivity Analysis for Discount Rate and Key Assumptions
The sensitivity analysis shown in Table 5-3 describes the bounds for the
net benefit estimates, using alternative discount rates and ranges of benefits
and costs. Part A of Table 5-3 shows the most likely estimates of net bene-
fits calculated with three different discount rates. Part B of Table 5-3 shows
the outcomes that would result for the worst case expected to occur by esti-
mating the cost savings at the lowest end of their range and the benefits fore-
gone at the highest end for three different discount rates. Part C of
Table 5-3 presents the estimates that correspond to the most optimistic case,
with cost savings at the highest end of their range and benefits forgone at
the lowest end of their range.
Step 3: Interpret Sensitivity Analysis
The sensitivity analysis shows that the net present value is sensitive to
the discount rate in that the magnitude of the net cost savings estimates vary
over a large range. However, changes in the discount rate alone over the
range employed are not enough to change the direction of the net benefit as-
sessment. In both the most likely and the worst cases the forgone benefits
exceed the cost savings for changing the use designation. Only in the most
optimistic case, where the cost savings are at the highest estimate and bene-
fits forgone are at their lowest do the cost savings exceed the forgone bene-
fits .
The recommendation that could be made from the sensitivity analysis is
that the change is likely to produce forgone benefits greater than the costs,
with only a small chance that the results would be otherwise. To the extent
there are effects that cannot be expressed in dollars, this range can also be
used to indirectly define what the dollar value of these effects would need to
be to change the evaluation. In the determination of the appropriate use
classification, the decisionmaker could then weigh this small chance.
5.3 DISPLAYING THE ASSESSMENT RESULTS
This section discusses three methods for displaying the results of a
benefit-cost assessment: narratives, arrays or matrices, and .graphical dis-
plays. Each method is described briefly, along with its advantages and dis-
advantages.
5-4
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Table 5-2. Sensitivity Analysis Calculations: Discount
Rate at 4 percent
Cost savings, $
Benefits forgone, $
Capital cost: 8.0 million in 1982
Range: 4.2 to 11.0
Operating cost: 1.5 per year
Range: None
Present value of operating costs:
32.2 million
Present value of total (most likely
case): 40.2
Range of present values of total
cost savings: 36.4 to 43.2 million
Recreation: 2.75 per year
Range: 0.9 to 5.75 per year
Present value: 59.1 million
Range: 19.3 to 123.5 million
Total (most likely case): 59.1
Range of total benefits forgone for
all cases: 19.3 to 123.5 million
Table 5-3. Sensitivity Analysis
Discount rate
Net present value
of cost saving minus
benefits forgone
(million $)
A. Most Likely Levels of Benefits Forgone and Cost Savings
2
4
6
-31.3
-18.9
-11.7
B. Cost Savings at Lowest Estimate—Benefits Forgone at Highest
2
4
6
-129.4
-87.1
-62.8
C. Cost Savings at Highest Estimate—Benefits Forgone at Lowest
2
4
6
29.9
23.9
20.5
5-5
U.S. EPA Headquarters Library
Mail code 3201
1200 Pennsylvania Avenue NW
Washington DC 20460
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Narratives
A narrative uses words to describe the results of an assessment. It can
be used to describe either qualitative or quantitative information and is a
simple, straightforward approach to displaying the results of an assessment.
Its main disadvantage arises when there are several types of beneficial and
detrimental effects to be weighed in the water quality decision. The evalua-
tion of these diverse benefits and costs may be aided by expressing some of
them with numerical estimates of benefits and costs. Combining narrative
information with the array method discussed below can ease the comparison of
benefits and costs. Examples of narratives used in combination with matrices
are shown throughout Chapter 6.
Arrays '
An array, or matrix, is a tabular display that contains written and nu-
merical descriptions of the outcomes of an assessment. Arrays are most ef-
fective when combined with the narrative display method discussed above.
An array organizes information in a simple yet visually effective manner. By
including qualitative information, the practitioner can easily describe the
nature of the benefits and costs and the degrees of confidence in the esti-
mates of either. If quantitative and monetized information is included, it can
be supplemented with descriptions of any benefits and costs that are not
monetized or cannot be quantified.
Arrays may be readily adapted to the wide variety of cases likely to be
encountered in an assessment of water quality programs. They do not
require assumptions about the relationships among the variables presented in
the array, and they make it easier for the practitioner to describe relation-
ships known only in various degrees of accuracy. Arrays are particularly
well-suited for displaying intangible benefits and costs and are used in both
Chapter 2 and Chapter 6 to highlight these issues. Most effective arrays are
used to organize information and to display substantive descriptions of the
information presented. Arrays that present too much information can be
divided into several arrays, but care is required to avoid unnecessary con-
fusion. Inadequate descriptions may be worse because they require that
users invest their own time.
Graphs
Graphs are an effective way of presenting information in a benefit-cost
assessment, but considerable caution and scrutiny are advised when using
them. Graphs can effectively show relationships between two variables, but
often the information required to draw them is simply not available. These
problems are less important for pie charts and bar graphs but are .prevalent
for tradeoff curves, which show relationships between two well-defined objec-
tives.
Figure 5-1 presents an example graph that can be used for two well-de-
fined objectives for a reservoir: flood control and recreation. Flood control
is measured on one axis in thousand acre-feet of water Impounded annually,
5-6
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
103 acre
impounded/yr
8
2 3
Recreation
103 visits/yr
Figure 5-1. Tradeoff curve.
and recreation is measured on the other in visits per year to the site.. In
this case, the shape of the curve shows that as more of the reservoir is used
for a single objective, larger and larger amounts of the other must be given
up.
The tradeoff curve shows a frontier of alternative combinations of flood
control and recreation that can be attained with a given reservoir in any year.
A point inside the frontier, like M, represents an inferior, or less attractive,
combination because more recreation can be attained while maintaining at least
the same amount of flood control. However, the tradeoff curve provides no
information about the relative attractiveness of points such as A through E.
All are oh the frontier, and society must choose which allocation of the two
objectives is most desirable. The extreme points A and E clearly show that
substantial amounts of either flood control or recreation must be given up for
an exclusive use. However, to know whether they are efficient from society's
viewpoint requires that the value of each good be known. The economic prin-
ciples behind the demand curve (which indicates that people will buy more as
price is lowered) imply that there may not be a simple one-to-one relation
between extra units of a commodity and the extra value from each unit.
While a useful concept, tradeoff curves are inappropriate in many bene-
fit-cost assessments. Many assessments for water quality policies will involve
more than two objectives that would be extremely difficult to express in a sim-
ple tradeoff relationship. For example, it is unlikely that suitable quantitative
units could be derived to meaningfully express such objectives as enhanced
ecological diversity, which may involve complex relationships that can easily
confuse the practitioner about tradeoff relationships.
5-7
-------
Key Elements
* Results of use attainability assessment
• Other background on river segment, water quality
1. Advanced treatment application
2. Previous standards documents
• List of benefits
• List of costs
• Data on benefits
- Primary
1 . Recreation survey— State
2. Recreation survey— U.S. and region
— Secondary
1. Published Federal, State, and local studies
2. Census
• Data on costs
— Primary
1. Engineering studies of treatment costs
2. Economic cost estimates
3. Financial cost estimates
— Secondary
1. Published Federal, State, and local studies
2. Financial/investment sources
• Techniques selected for measuring benefits
1. Direct survey
2. Participation survey
3. Travel cost
4. Hedonic
• Techniques selected for measuring costs
1 . Engineering estimates
2. Computer-assisted model ing techniques
3. Economic cost estimates
• List of key assumptions in measuring
benefits and costs
• Discount rate
• Time horizon
• Sensitivity analysis
• Present value of net benefits -
• Assessment of distribution
• Cost impact assessment
• Final assessment document
( .. _.....
Status of Data for Assessment
Not
Applicable
Available
Requested
Unavailable
•v
Figure 5-2. Checklist for a water quality standards benefit-cost assessment.
5-8
-------
In summary, tradeoff curves are convenient display techniques but are
inappropriate when the objectives are not well defined or easily expressed in
quantitative or monetized units. The practitioner should use this technique
with a great deal of caution and only when substantial data are available.
5.4 BENEFIT-COST ASSESSMENT CHECKLIST
This section describes a checklist that can be used in organizing a bene-
fit-cost assessment. The. example is for a water quality standards, case, but
the checklist is a general one.
The checklist in Figure 5-2 provides a means, of tracking the various
steps involved in the assessment process. This checklist can be tailored to
fit the needs of each assessment by varying the types of benefits and costs
included and by using the columns at the right-hand side of the list. These
columns allow the checklist to accommodate the range from simple qualitative
to complex quantitative assessments by designating the status of the data
required for the assessment. A qualitative assessment can be conducted even
with such status categories as "not applicable" and "unavailable." The com-
plex assessment is made easier when the categories are "available" or
"requested." The larger the "unavailable" category, the more strongly a State
should consider using additional resources, either inside or outside State
government, to obtain the needed information, especially if benefits or costs
are likely to be sizable and if the decision is unclear after a qualitative
.assessment. For a complex case involving potentially large benefits and costs,
the extra value of acquiring the necessary data can be substantial. Data on
a specific water body and the use of the techniques described in this hand-
book can greatly simplify the complex quantitative assessment and assist in
evaluating a proposed water quality action.
5.5 SUMMARY
• • A sensitivity analysis for key variables and assumptions
employed in an assessment is essential for a plausible assess-
ment.
Sensitivity establishes a range of outcomes possible for the
assessment and will show when the assessment is sensitive to
its assumptions.
Three key components of an assessment that will be essential
in the sensitivity analysis are the discount rate and estimates
of benefits and costs.
A narrative describes the results of an assessment in words.
Its main advantage is its simple straightforward nature, while
its disadvantage is in presenting results for more .complex
assessments.
Arrays or matrices are tabular displays that contain written
and numerical descriptions. Narratives are often combined
5-9
-------
with the arrays as an effective display technique and are
suited for a wide range of assessments.
Graphs can effectively illustrate relationships between objec-
tives that are quantified or monetized. However, the informa-
tion necessary for these display techniques is frequently un-
available.
Tradeoff curves show a frontier of alternative combinations of
quantities for two objectives. Caution and scrutiny are ad-
vised in using these curves in an' assessment because the
necessary information is often unavailable or it is impossible to
apply to water bodies with a wide range of uses.
A checklist is one way of organizing the procedures and infor-
mation in a benefit-cost assessment.
5-10
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
CHAPTER 6
BENEFIT-COST--SAMPLE SCENARIOS
6.1 INTRODUCTION
In water quality programs, the benefit-cost assessment practitioner must
evaluate a wide range of policy actions. These could include both minor
changes in designated uses for an intermittent stream and major changes .in
designated uses or in advanced treatment or combined sewer overflow to
provide recreation and intangible benefits that require substantial investments
by cities and firms. Thus, the framework for evaluating these policy actions
must be capable of comparing Incremental benefits and costs for a diversity of
cases yet provide for a consistency in the application and presentation of the
assessment. Even though scenarios are presented only for potential water
quality standards decisions, the range of issues is broad enough to be useful
to other water quality programs.
To illustrate the type of assessments that might arise in water quality
programs, this chapter develops three sample scenarios—simple, medium, and
complex. Each scenario is designed to build on the preceding one, as new
dimensions are added. Each scenario is introduced by a brief description of
hypothetical sample cases. Each scenario refers the reader to the relevant
handbook chapter(s) that provide more detailed discussion on specific issues.
Although the step-by-step framework illustrated in the following scena-
rios is both systematic and flexible enough to accommodate most of the benefit-
cost assessment needs of various water quality programs, the values it assigns
to benefits and costs should be regarded as approximations rather than abso-
lutes. This note of caution has nothing to do with the framework itself, but
with the specific information--e.g., poor quality data on linkages between
water quality, fish propagation, and recreational fishing—used to estimate
either costs or benefits.
The following sections of this chapter present the three scenarios for
benefit-cost assessment. Specifically, Section 6.2 describes a simple case
scenario that uses only qualitative information, Section 6.3 presents the medi-
um case scenario, providing some quantitative benefits and costs, and Sec-
tion 6.4 contains a complex case scenario involving multiple benefits and costs.
Finally, Section 6.5.summarizes the key issues in the scenarios.
6-1
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
6.2 SIMPLE CASE SCENARIO
Introduction
The strength of benefit-cost assessment lies in its ability to organize ma-
terial in a consistent manner yet remain flexible enough to accommodate a wide
range of cases. This simple scenario illustrates a qualitative benefit-cost as-
sessment and demonstrates both the consistency and flexibility in this evalua-
tion procedure.
In many water quality applications, the potential benefits and costs do
not justify anything more than a qualitative benefit-cost assessment. Although
a qualitative assessment does not quantify the information it organizes, it does
provide a framework for presenting the character of the individual problem
and for describing the judgment employed to make the assessment. For
example, water quality standards decisions where qualitative assessment may
be appropriate include stream-specific standards, such as the following:
Public water supply designations for streams that have never
been so used or that—because of low cost alternatives—are not
likely to be so used in the future. Possible changes include
removing the water supply use and adding recreation or agri-
cultural uses that the water supply use might have precluded.
Primary contact recreation uses for a stream that currently has
few access points, as well as water quality limitations. Pos-
sible changes include limiting types of recreation or changing
uses to accommodate agricultural and industrial activities.
Simple Case Scenario Format
I. Define the Action
.A* State is reviewing the designated uses for a specific segment of a
river currently designated for use as a public water supply, although it has
never been used in this capacity, and for primary body contact recreation.
The action to be assessed is the removal of the water supply use and the
addition of an irrigation use. The segment is located in a portion of the
State that produces a substantial amount of agricultural products. The seg-
ment is a primary source of water recreation in the area and supports fish-
ing, swimming, and limited boating.
II. Translate the Effects into Beneficial Outcomes and Costs
Although the use change will cause slightly lower levels of dissolved
oxygen and small increases in the levels of several other biological and chemi-
cal water quality parameters, recreation activities will be unaffected. The
primary benefit will be the increase in high-quality water available for irriga-
6-2
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
tion, resulting in cost savings for farmers who presently have irrigated farm-
land or who will be able to irrigate additional farmland.*
In terms of forgone benefits, the cost of the action is the loss of a poten-
tial water supply source. However, groundwater sources exist, and residents
are presently using them at lower costs and are expected to continue their
use in the future. Nothing (e.g., shortages or contamination) is expected to
threaten these groundwater supplies. .
III. Calculate the Value of the Beneficial Effects Based on Willingness to Pay
The value of the benefits of the change would equal the farmers' willing-
ness to pay to obtain the water for irrigation rather than do without it. The
benefits are expected to be positive on this basis, but not necessarily large
because some alternative irrigation supplies exist.
IV. .Calculate the Value of the Detrimental Effects Based on Opportunity Costs
The opportunity cost of this action is loss of the potential source of
drinking water. The alternative supplies of drinking water make this oppor-
tunity cost near zero. If future demand and supply of drinking water should
change substantially, the State could reconsider the designated uses because
the long-term physical effects on the water will be limited in nature.
V. Compare the Total Benefits and Costs
The benefits of the action are the amounts the farmers would be willing
to pay from this new irrigation source to irrigate their farmland without the
action. The costs of the action, however, are zero, because they consist
entirely of the lost opportunity for an alternative supply of drinking water,
which can be obtained from existing groundwater resources. Thus, the total
benefits of the action are greater than the costs. Therefore, the action will
provide positive net benefits for the State and country. Recreation activity
will be maintained, and cost savings for farmers will be greater than the loss
of a potential source of drinking water.
VI. Assess the Plausibility of the Results
The results are not likely to change under almost all possible circum-
stances. Only if dramatic changes occur in the drinking water situation--
i.e., if groundwater supplies become contaminated or suddenly in short
supply-would the outcome be different.
*This particular segment has ample flow but farmers have not been per-
mitted to use if for irrigation. Withdrawal for irrigation will not noticeably
affect recreation activities or fish populations.
6-3
-------
6.3 MEDIUM CASE SCENARIO
Introduction
As a group, medium cases can be distinguished from simple and complex
cases. While simple cases usually require only a qualitative assessment of
.benefits and costs, medium cases normally involve some quantitative measures
of benefits and costs. Further, while medium cases usually involve only one
type of benefit—such as recreation—and the application of measurement ap-
proaches and results from other studies, complex cases often require develop-
ment of case-specific measurement approaches.
One practical way to distinguish between simple and medium cases is to
conduct the qualitative assessment and then to judge whether it provides
enough information for a clearcut decision. If ;the outcome is not clear, the
more quantitative medium case assessment may be needed. Water quality
standards decisions that might require medium assessments include the follow-
ing:
Fish and wildlife propagation use for a stream that is not
being attained because of industrial dischargers, the treat-
ment options are limited to high-cost land-treatment solutions,
and the river is not an important recreation source. Change
could be to provide for a less restrictive fish arid wildlife use
or for agricultural/industrial use.
Fish and wildlife propagation use for a major river tributary
with industrial dischargers. A change in designated use to
accommodate existing industrial cooling would maintain rela-
tively low levels of dissolved oxygen in a large segment of the
river, primarily affecting fishing. The assessment would com-
pare the loss in potential fishing benefits with cost savings to
industrial firms.
Fish and wildlife propagation use for a stream whose hydro-
logical equilibrium has been affected by irrigation. A change
in designated use to accommodate agricultural irrigation would
cause the loss of a potential warm water fishery in the stream.
The assessment would compare loss in potential fishing benefits
with cost savings to agricultural irrigation users.
Attaining a limited warmwater fishery (sunfish, carp, catfish)
use for a channelized stream would require an advanced treat-
ment plant for municipal wastes. One change would be to
provide roughfish passage. The assessment would compare the
loss in potential fishing benefits with the cost of advanced
treatment.
6-4
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Medium Case Scenario Format . .
I. Define the Action
A State is reviewing the designated uses for a specific segment of a
river currently designated for agricultural arid industrial use and secondary
recreation, primarily boating. The action to be evaluated is a use change to
provide a warmwater fishery in addition to other uses. The river segment is
capable of sustaining a warmwater fishery if discharges .from a municipal
treatment plant are reduced to improve the level of dissolved oxygen.
II. Translate the Effects into Beneficial Outcomes and Costs
The effects of the change will be an improvement in dissolved oxygen
levels to sustain the warmwater fishery under all flow conditions. This
change will provide a new source for smallmouth bass recreational fishing
while maintaining the existing uses. The primary cost of the change will be
additional treatment of municipal wastes, but advanced treatment is not antici-
pated.
Ml. Calculate the Value of the Beneficial Effects Based on Willingness to Pay
A travel cost model can be used to estimate the recreational fishing bene-
fits. These benefits are estimated to be $70,000 per year and are assumed to
remain at that level for the next 20 years.* This amount represents fisher-
men's willingness to pay for the water quality improvement, but does not meas-
ure nonuser values or intangible benefits associated with the water quality
improvement.
IV. Calculate the Value of the Detrimental Effects Based on Opportunity Cost
The municipal discharger will be required to upgrade the quality of its
treatment plant. From engineering estimates, this upgrading is expected to
require an initial investment of $500,000 but no increases in operating ex-
penses. There Will be a small economic impact on households who pay for the
treatment services. (See Chapter 4 for details on measuring costs.)
V. Compare the Total Benefits and Costs
To compare them, benefits and. costs must be converted into present
values. Costs incurred in the present year are already in present value
terms ($500,000), but the benefits must be converted into a present value
equivalent from a stream of annual dollars over 20 years. Using a social rate
of time preference as presented in Chapter 2, these benefits are discounted
*See the complex scenario for the use of a travel cost model to estimate
recreational fishing benefits. The results are summarized for the medium case
to minimize duplication among the scenarios. See Chapter 3 for general dis-
cussion of the travel cost approach.
6-5
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
at 2 percent for a total 20-year value of $1,148,000, Thus, total benefits
resulting from the change are $1,148,000, and total costs are $500,000, result-
ing in a positive net benefit of $648,000. Adding the new use will therefore
provide benefits in excess of the required investment costs.
VI. Assess the Plausibility of the Results
The key element influencing the sensitivity of this scenario is the selec-
tion of the discount rate. To determine the plausibility of the results, net
benefits should be calculated with different discount rates and compared.
Chapter 2 recommended a high estimate of 5 percent for the social rate of time
preference, and the Office of Management and Budget (OMB) requires the use
of 10 percent for major regulations. Net .benefits calculated with these rates
are listed below. .
Discount rate
(percent)
2
5
10
Net benefits
($)
648,000
372,340
95,980
As indicated, the use change would .produce positive net benefits under any
of the three discount rates, implying that benefit estimates are not sensitive
to the discount rate selected. Sensitivity analyses also may be performed for
the estimates of benefits and costs--for example, with benefits and costs esti-
mated at ±30 percent of the average value. More details on using sensitivity
analysis are provided in Chapter 5.
In assessing the plausibility of the results one might .also consider the
distribution effects discussed in Chapter 2. In this sample case these effects
are an inconsequential part of the assessment because no one group is adv-
ersely affected. However, distribution effects vary from case to case and
should be addressed.*
6.4 COMPLEX CASE SCENARIO
Introduction
The complex case scenario is distinguished from the simple and medium
case scenarios by several characteristics, including its consideration of multi-
ple types of tangible benefits, intangible benefits, and investment in waste
treatment beyond the technology-based requirements. In addition, benefits
and costs in complex cases are likely to be an order of magnitude, or more,
above those in medium cases. In fact, if benefits and costs are not signifj-
*The complex scenario that follows provides a detailed assessment of dis-
tribution effects.
6-6
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
cantly greater than those of the typical medium case scenario, use of the com-
plex case scenario is not recommended because the effort it requires would
not be justified. .
Measurement and valuation of benefits and costs may require several
methods discussed in Chapters 3 and 4. The types .of benefits and costs, data
availability, and available staff resources will influence the level of detail in
the assessment. For controversial decisions, a State may decide to obtain
assistance from outside the agency to conduct the assessment.
Complex cases may include the following:
Adding fish and wildlife propagation and primary contact
recreation use designations for a stream that will require
advanced treatment for municipal discharges. Benefits will
include multiple types of recreation activities, intangible
benefits such as enhanced ecological diversity, and nonuser
benefits. Costs will include .investments by firms in waste
treatment beyond investment required for technology-based
requirements.
Providing public water supply from a stream that will require
advanced treatment for municipal wastes. Water supply bene-
fits could be for more than one downstream municipality.
Complex Case Scenario Format
I. Define the Action
A State is reviewing the designated uses for a 10-mile segment of
River 1, located between Cities A and B. The river is currently designated
for secondary contact recreation, but the State is considering adding swimming
and fish and wildlife propagation. These uses cannot be attained by imposing
the technologies required under the Clean Water Act (Section 30i(b)(2)), but
they could be attained with more stringent controls on municipal and industrial
point dischargers.
River 1 drains an area of 7,386 square miles, and most of its length is
characterized by steep banks and rugged terrain. The major point sources of
discharges on the river segment of interest are iron and steel facilities and a
municipal sewage discharge just outside City B and upstream from City A.
The river is navigable the entire year, and a considerable amount of coal
is barged through several locks and dams operated by the U.S. Army Corps
of Engineers. River 1 is currently used for boating recreation and for activ-
ities near water at several parks located along its banks. The most notable
water quality problems limiting swimming and fish and wildlife propagation
have been associated with the dissolved oxygen and ammonia levels, with fre-
quent violations of current standards at low flow.
6-7
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
II.
Translate the Effects into Incremental Beneficial Outcomes, Costs, and
.Economic Impacts
A. Benefits
Recreational fishing
Swimming
Activities "near water"
Navigation
Ecological diversity
Nonuser/nonuse benefits.
B. Costs
Capital, and operating costs for steel plants
Capital and operating cost for City B
C. Distributional Impacts
Reduced profitability of steel plants
Effects on operations (will shutdowns occur?)
Increased sewer charges for municipal taxpayers
Employment
Price
Impact on firms within industry
III. Calculate the Value of Beneficial Outcomes Based on Willingness to Pay
•A. Recreational Boating Benefits
Step 1
Determine what data are available for recreation on or near River 1.
Likely sources include the State recreation plan, the EPA 208 management
plan, the 1977 Department of Interior outdoor recreation survey, and the
U.S. Army Corps of Engineers.
Step 2
Use the Corps of Engineers recreation survey data on River 1,
which contains information on users and where they are from. This informa-
tion will allow estimation of recreation benefits with the simplest version of
the travel cost model. (See Chapter 3 for more details on travel cost
models.)
A number of implicit assumptions are used in this scenario. For ex-
ample, the State is assumed to have the information necessary to estimate the
current and potential demand for fishing and swimming for River 1. The
simplest measure of travel cost is used even though it excludes the cost of
travel time, time spent onsite, and the influence of substitute sites on the
demand for River 1 recreation. Also assume that the travel cost to the site
6-8
-------
is capable of capturing all the factors that influence the decision to fish or
swim along the river. This implies that no changes in access, docks, and
other site features are occurring. (See the case study of the travel cost
model in Chapter 3.)
Equation (6.1) illustrates the travel cost demand equation for these as-
sumptions. The equation omits several origin-specific variables (such as in-
come, age, education, etc.) that determine recreational demands for a site be-
cause they would only change the position of the demand curve for each
origin zone. The subscript i refers to these omitted variables:
A)
(6.1)
where
A = activities that would be permitted by the water quality use classi-
fications (A = B for beatable, F for fishable, S for swimmable),
V. = number of visits to the site from origin zone i, and
POP. = population of origin zone i.
Step 3
Calculate recreation benefits for existing water quality level. These
calculations are needed to obtain the baseline and must be netted out from the
benefits of the additional use designations. Use the data in Table 6-1 to esti-
mate travel cost model. Remember: these are user benefits only.
Table 6-1. Demand for Recreation for River 1—Water Quality
at Level Suitable for Boating
Limits
of zone
of origin
(miles)
1 to 20
21 to 40
41 to 60
61 to 80
81 to 100
100+
'1980
population
(1,000)
1,000
1,500
2,000
2,500
3,500
5,000
Total No.
party visits
1,650,000
600,000
200,000
50,000
25,000
2,500
Visits per
1 ,000
population
1,650
400
100
20
7
0.5
Consumer
surplus per
individual
($)
2.40
1.80
1.40
1.00
0.60
0.20
6-9
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Calculation of Benefits
n
Benefits =
TO*
, .'/VP.'
TC -
B)dp.
(6.2)
Benefits = (number of individuals) x (benefits per individual)
where
TC* = the travel cost at which there would be a zero visit rate, and
TC. = travel cost for ith origin zone.
Figure.6-1 illustrates the consumer surplus estimates that result from
this calculation. Two assumptions are implicit for the estimation. First, as
with XYZ in Figure 6-1, calculate the representative individual's consumer sur-
plus in an origin zone, and then assume that all individuals in the origin zone
have the same consumer surplus. This procedure is possible because use
(i.e., the visit measure) is considered to arise from the population as a
whole. (The calculation would be different for travel cost models estimated
from survey data. See Chapter 3 for specifics.)
Travel
Cost
$/visit
Consumer
surplus
fi(Pi,B>
Pop
Figure 6-1. Travel cost demand function and
consumer surplus with boatable water.
For example, in Table 6-1, the fifth column reports an example of the
estimated individual consumer surplus by origin zone. Multiplying each con-
sumer surplus times the population and adding across origin zones gives the
aggregate consumer surplus for the site with boatable water:
6-10
-------
1
1
1
^^p
1
1
•
1
1
1
1
•
1
•P_
1
1
••
1
M
\
6
Benefits for Boatable Water, 1980-81 = \ population x (consumer surplus
^•4 per individual)
Zone 1
A-
= '$(1,0.00,000 x 2.
Zone 4
+ ' 2,500,000 x 1
= $(2,400,000 + 2,
+ 2,100,000 + 1,
Benefits for Boatable Water =
B. Recreational Fishing
Step 1
Use the results on
(see travel cost case study,
zone i=1
Zone 2 Zone 3
A *.
40' +f 1,500, 000 x 1.80" +'2, 000, 000 x 1.40'
Zone 5 Zone 6
.00"+ 3,500,000 x 0.60' +'5,000,000 x 0.20)*
700,000 + 2,800,000 + 2,500,000
000,000)
$13.5 million for the site with current water
quality levels.
Benefits
fishing benefits in the Monongahela River study
Chapter 3, for details). The results from the
Moripngahela study can be used to calculate benefits analogous to those re-
ported in Table 6-2. Note:
determining the plausibility
River 1 and the Monongahela
practitioners might consider three questions in
of results: What similarities exist between
River? What are the problems in adapting the
results? How significant are they?
Table 6-2. Demand for Recreation for River 1 --Water Quality
at Level Where
Limits
of zone 1980
of origin population
(miles) (1,000)
1 to 20 1,000
21 to 40 1,500
41 to 60 2,000
61 to 80 2,500
81 to 100 3,500
100+ 5,000
a
Gamefish (Bass) Can Live in River
Consumer
Visits per surplus per
Total No. 1,000 individual
party visits population ($)
3,400,000 3,400 0.10
1,500,000 1,000 0.10
800,000 400 0.10
500,000 200 0.10
350,000 100 0.10
50,000 10 0.10
These calculations assume a parallel shift in demand. It need not be parallel
in particular applications (see
travel cost case study, Chapter* 3).
6-11
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Step 2 .
For this framework, the travel cost model provides the relationship
between the demand for the recreation site and the water quality use designa-
tion. Thus, an improvement in water quality from boatable (B) to fish-
able (F) can be expected to shift the demand for the site's services. This
conclusion follows from the. site's ability to support both types of recreational
activities under the higher designated use. Figure 6-2 illustrates the type of
shift involved.
^Change in consumer
surplus with
fishable water
Figure 6-2. Travel cost demand function and change
in consumer surplus with fishable water.
To calculate the incremental benefits associated with this change in use
designation, we need to estimate YXWV. In terms of the travel cost demand
model, YXWV would be given as:
Benefits (incremental to fishable) =
n -
E P°Pi7
TC**
TC*.
F)dp. -
TC .
5=1
/"
j J
TC.
B)dp.
(6.3)
The second term in Equation (6.3) simply repeats the benefit estimate
for the site with the existing designation which allows boating. Thus, the
increment to benefits because of the change in water quality is being calcu-
lated. This example assumes that individuals from the same origin zones are
using the site, and no users from new origin zones so Equation (6.3) can be
rewritten as the sum of the increments to the individual benefits realized in
each origin zone, as in Equation (6.4):
6-12
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Benefits for
increment
to
fishabfe
water
n
= E popi
'TC**
/'
TC*i
f. (p., F)dp. - J f.(Pi, B)dP]
TC
(6.4)
number of
individuals
increment to each individual's
benefits from use designation
For the representative individual, this increment is YXWV in Figure 6-2.
Table 6-2 reports some illustrative increments in the fifth column, and
the calculation process is then similar to that described above,;
Benefits for Increment
to Fishable Water, 1980-1981
Zone 1
visits x travel costs
Zone 2
Zong 3
= $(1,000,000 x 0.10 +1,500,000 x 0.10 + 2,000,000 x 0.10
Zone 4 Zone 5 Zone 6
__ _ _ .^^^^^^^^^^^^^^^^^^^fc. .^^^^^^^^^^^^^^^^•••••k
'2,500,000 x 0.10 V 3,500,000 x 0.10V5,000,000 x 0.10 x
= $(100,000 + 150,000 + 200,000 + 250,000 +
350,000 + 500,000)
Benefits for increment .to fishable water = $1.6 million per year.
C. Swimming Benefits
Step 1
Use the results on swimming benefits in the Monongahela River
study (see travel cost case study, Chapter 3, for details) to calculate infor-
mation such as that in Table 6-2.
Step 2
To calculate the benefits from swimming use, the frame of reference
must be defined. Specifically, which incremental benefits are of more
interest — those associated with moving from a fishable to a swimmable use des-
ignation, or those associated with moving from boatable to swimmable? The
method used to estimate these incremental benefits will depend on the refer-
ence point. For example, in Figure 6-3, which shows all three demand func-
tions, movement from fishable to swimmable leads to incremental benefits (per
individual) of VWUT. A change from boatable to swimmable includes this in-
crement along with the increment associated with the improvement to fishable
(YXWV).
6-13
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
^Change in consumer
surplus with
swimmable water
V
Pop
Figure 6-3. Travel cost demand function and change
in consumer surplus with swimmable water.
The calculations would follow the same outline used in the preceding two
examples. Important in this case is the point of reference used to calculate
the incremental benefits. If procedures from previous section are followed:
Benefits for the increment from fishable to swimmable water = $750,000.
D. , Recreational Benefits of Activities Near Water
Improvements in water quality will make additional acres usable for
activities .near River 1, including, for example, hunting, birdwatching, hik-
ing, photography, and sightseeing. This example shows what can be done to
estimate the benefits when demand information is not available. In particular,
a participation model can be used to predict the increase in recreation activ-
ities near water and a recreation day value estimate can be used from other
sources. (See participation survey case study, Chapter 3, for details.)
An important issue that arises with this approach is the consistency of
benefits derived from the travel cost model with those from the participation
model. Specifically, the travel cost model describes the demand for recrea-
tional site services which can be used in a variety of recreational activities,
including those "near water." Since the participation model for predicting
demand for recreation near water may be measuring benefits reflected in the
travel cost model, some double counting can occur.
The following five steps can be used to estimate the value of the in-
crease in near-water activities.
Step 1
Determine the availability of data on activities near water for
River 1.
6-14
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Step 2
Estimate the increase in near-water activities resulting from the fish
and wildlife propagation use designation. Analysis of River 1 determined that
8,931 acres along the river are currently suitable for these activities.
Achieving the use designation was predicted to make another 2,700 acres suit-
able for near-water activities and to improve the recreational facility rating
by 1 unit (on a scale of 1 to 5).
Step 3
Estimate the change in the probability of participation (AP) in the
near-water activities among the population as a result of the improvements.
Using a model (see Chapter 3, participation survey case study for details)
that correlates participation in near-water activities with available water recre-
ation area and a rating of the recreational facilities, the increased probability
of participation can be calculated as:
AP = 0.38485 /change in acreage of recreationY+ 0.03142/change in recrea- ^
I water available per capita j \ tional facility rating
\
AP = 0.38485
AP = 0.0319.
2,700 new acres
12,000,000 population in
Cities A and B)
+ 0.03142 /I unit improvement
I recreational facility
\rating
Step 4
Estimate the change in participation days attributable to River 1
between Cities A and B. Specifically, using the national average of 2.0 near-
water activity days per participant per year, additional near-water activity
days can be calculated as: •
Change in near-
water activity days
/change in \ / Regional\
I probability \ [population \
I of. near-water
yparticipation
(0.03I9)
I Cities A I
\and B /
x (2,000,000)
'Near-water'
days per
participant
(2.0)
Change in near-water activity days = 127,600/year.
Step 5
Estimate the value of additional near-water activity days. Re-
searchers have estimated values per near-water activity day .to range from
$12.00 to $18.50 (Volume II will provide more details on these values). Multi-
plying these values by the number of additional near-water activity days
gives the following range of values for the additional activities:
6-15
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
(12.00/near-water activity day)(127,600 days) = $1.5 million
(18.50/near-water activity day)(127,600 days) = $2.4 million
Note: Remember the limitations of this approach:
1. The valuation step is not based on willingness to pay.
2. Limitations are caused by transferring the model from one study to
another.
E- Benefits of Improved Ecological Diversity
Attaining the fish and wildlife .propagation use designation also
means enhanced ecological diversity of River 1 between Cities A and B.
These benefits are not quantifiable at the present time but would include more
diverse flora .and fauna, enhanced diversity of fish species, and other related
elements. Note: These benefits will be listed in the policy array as "positive
nonquantifiable" (see Chapter 2 on intangible benefits), allowing the decision-
makers to better focus on these questions. In some cases, in fact, it may be
possible to list specific species that the new use will add or preserve.
F, Benefits of Improved Navigation
Although achieving the .designated use may also benefit navigation
in River 1, indications are that the magnitude of these benefits is negligible.
Thus, no attempt will be made to quantify them (see Chapter 3 on measuring
benefits to firms).
G. Nonuser/Nonuse Benefits
In addition to recreation in or near the water by current users, in-
trinsic benefits—option values, existence values/ and aesthetics--may account
for important benefits of improved water quality. Option value refers to the
value current and potential users place on having the option to use the water
resource at some future time. This value is assigned to the resource because
there is some uncertainty regarding its future availability or regarding future
demand. Existence benefits are measured by the value people place on ac-
tions that ensure a resource is (or will be) there, regardless of their actual
or potential use. This is sometimes termed vicarious enjoyment or attributed
to a bequest motive. Aesthetics refer to beauty that may be appreciated by
users and/or those who reside or travel nearby.
Many unresolved issues exist concerning the inclusion and valuation of
intrinisic benefits. For example, there is not yet any agreement oh whether
aesthetics should be included in intrinsic benefits or measured separately or
indeed, whether they be measured at all. Thus, intrinsic benefits are in-
cluded in this assessment as "positive not quantified" (see the contingent
valuation case study in Chapter 3, which highlights an approach for quantify-
ing and monetizing these benefits,)
6-16
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
H. Estimate Benefits That Will Accrue In Future Years
Assume that River 1 recreation benefits will remain constant for the
25-year lifetime used for the analysis. This is probably a conservative as-
sumption because, with visit rates staying the same, any population growth in
the area would lead to more recreation (unless congestion becomes a problem).
This growth is offset, however, by other important factors such as the in-
fluence of substitute sites that are excluded from the estimation models.
Thus, each benefit estimate can be considered to be a constant stream over
the project lifetime.
IV. Value the Detrimental Effects Based on Opportunity Costs and Calculate
Economic Impacts
A. Calculate Increased Capital and Operating Costs for Model Steel
Plant
Step 1
Determine availability of data sources (see Chapter 4 for details on
likely sources of cost data).
Step 2
Use the model plant technique discussed in Chapter 4 along with
the wasteload allocation to estimate the capital costs of achieving the fish and
wildlife propagation use designation. A related study of pollution controls for
steel plants suggests that the technologies shown in Table 6-3 will be needed
to meet the regulation. Two steel plants correspond closely with the model
plants; two are about 3/4 the size assumed for the plant; and 1 plant is 1-1/4
times larger than the model plant. Table 6-4 shows capital cost estimates for
these plants.
Table 6-3, Capital Equipment and Cost for Model Plant
to Meet Regulation
Capital equipment
Wastewater discharge treatment pond
Water integrated flow system
Pretreatment and handling system
5 segments 30" piping
Total
Cost (million $)
1.2
1.6
1.2
2.0
6.0
Step 3: Calculate Operating Costs
Assume for the model plant analysis that operating costs are
roughly 0.05 of capital costs per year for the life of the plant, regardless of
the size of the plant. Using this estimate and the capital costs from Table
6r4, calculate:
6-17
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Table 6-4. Estimated Capital Costs
Number of. plants
Total
2
2
I
5
Percent of model plant
0.75
1.00
1.25
Total estimated capital
cost (million $)
9
12
7.5
28.5
Operating costs/year = 0.05 x $28.5 million = $1.4 million/year.
The problem of converting costs into comparable measures is addressed in
Chapter 4.
B. Calculation of Capital and Operating Costs for City B
To meet the requirements of the designated use of fish and wildlife
propagation, City B will have to add an advanced waste treatment plant for
its sewage wastes. Capital and operating costs can be calculated using the
model plant method for advanced waste treatment (see Chapter 4).
Step 1
Determine availability of data on costs of advanced treatment plants
(see Chapter 4 for details on availability of data on advanced treatment
plants).
Step 2
Determine additional capacity needed for the advanced waste treat-
ment plant. Assume that the facility is to provide for.maximum daily flow
equal to two times expected average daily flow. (This information would come
from a City B sewer study, for example.) The additional advanced waste
treatment capacity required is 60 Mgal/d.
Step 3:
Table 6-5 shows the capital cost estimate resulting from the model
plant approach.
Table 6-5. Cost of Treatment Plant, City B
Plant capacity
Capital cost
60 Mgal/d
$5 million
6-18
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Step 4
Use the model plant to estimate operating costs. Based on other
studies 10 percent of capital costs can be used to represent operating costs
for City B:
Operating cost = Capital cost x operating cost factor
$500,000/yr = 5 million x 0.10.
C. Perform Plant Closure Analysis to Assess Impact of Regulation
on Plant Operations
The plant closure test provides a straightforward assessment of the
impact of the regulation following the rules of thumb referred to in Chapter 2
for plant closures. There is considerable uncertainty involved in actual plant
closure decisions, so this appraisal will give a benchmark rather than a com-
.plete determination of plant closures.
Step 1
Determine data availability (see Chapter 2 and Chapter 4 on finan-
cial data sources that are needed for this test).
Step 2
Determine opportunity cost of capital and average liquidation value
to the two steel companies that own the five steel plants along River 1. Use
these estimates in the following formula to determine the critical rate of
return for closures:
Average liquidation value (m) x opportunity cost of capital (r)
= critical rate of return for closure
American Steel: m x r = critical rate of return
0.50 x 5 percent =2.5 percent.
Riverton Steel:
Step 3
m x r - critical rate of return
0.60 x 6 percent =3.6 percent.
Compare critical rate of return with plant's rate of return on re-
placement cost with the regulation to determine potential Closure candidates:
6-19
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
American Steel: Rate of return vs. critical rate of return
5.0 percent vs. 2.5 percent.
Riverton Steel; Rate of return vs. critical rate of return
4.8 percent vs. 3.6 percent.
Based on this simplified test using rules of thumb, there are no closure
impacts. If closure were indicated, .more detailed calculations and data would
be needed to assess the issue more thoroughly. Volume II will address these
more detailed comparisons.
D. Perform Profit Reduction Test
Another test to assess the magnitude of the impact of the regulation
on companies is the profit reduction test. A simplified version of this test is
illustrated below.
Step 1
Determine available data (see Chapter 2).
Step 2
Estimate reduction of profits as a percent of current dollar returns
for the two steel companies:
American Steel: 5 percent reduction in current dollar returns.
Riverton Steel: 8 percent reduction in current dollar returns.
These reductions are both relatively small impacts, so no additional calcula-
tions are needed.
E. Determine Impact on Households in City B
Another issue in the impact assessment is who bears the cost of ad-
vanced treatment. Issues that will need resolution include the following:
Does City B receive EPA assistance for the advanced treatment
plant? If so, the citizens of City B and all U.S. taxpayers bear
the cost impacts. If not, only residents of City B bear them.
Will future costs be incurred by the residents of City B when these
costs are passed along?
For purposes of illustration, no advanced treatment assistance from EPA is
assumed.
6-20
-------
Step 1
Determine available data (see Chapter 2 on municipal waste treat-
ment cost impacts for more on available data).
Step 2
Determine impact of increased cost on residents of City B. Sewage
costs will increase an average of $15 per year per household for residents of
City B.
V, Compare Total Benefits and Costs
A. Check for Less Than Fully Employed Resources
Volume II will present guidelines on determining whether money
costs overstate true social costs due to the presence of significant unemployed
resources. Assume the review shows that unemployment equals 5 percent in
both City A and City 'B and that a major percentage of materials comes from
outside the area. There are no problems with overstating costs in this case.
B. Discounting Benefits and Costs
This is a crucial step in the process of assembling the benefit-cost
assessment. There are several key steps to. be performed (see Chapter 2 for
detailed treatment of these steps).
Step 1
Select discount rate for converting future benefits and costs into
present values. OMB guidelines for regulatory impact analyses recommend a
real discount rate of 10 percent. This seems high for a real rate that re-
flects either the opportunity cost of capital or society's preferred rate.
Thus, a range of rates should be used, with 10 percent on the upper end,
and the sensitivity of results should be compared to the discount rate. A
social rate of time preference procedure suggested in Chapter 2 is illustrated
in Step 2 below.
Step 2
Discount annual monetized benefits into present values.
Annual Monetized Benefits
Fishing
Swimming
Near water
Total
million $
1.55
0.75
1.5 to 2.4
3.8 to 4.7
6-21
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Since the benefit amounts are constant across time periods, to convert mone-
tized annual benefits into present value equivalents, use the following formu-
la. It calculates the present value (P) of a stream of annual benefits (A) for
n years (assumed project life is 25 years) and interest rate (i):
1 -
1
P = A
(1 * i)
n
Table 6-6 shows the present values of benefits calculated with discount rates
ranging from 2 to 10 percent. Monetized benefits range from roughly 35 to
92 million 1981 dollars.
Table 6-6. Present Value of Benefits
Discount rate
(percent) .
Present value
of benefits
(million $)
2
4
6
10
74.2 to 91.8
59.4 to 73.4
48.6 to 60.1
•34.5 to 42.7
Step 3
To simplify matters, assume that all the capital costs are incurred
in 1981, so that they are already present values. Based on the Lind proce-
dure, these displaced private investments by the steel companies are multi-
plied by the shadow price of capital--1.9. The capital costs for City B treat-
ment plant are assumed to displace consumption (so they do not need adjust-
ing by the shadow price of capitai).
The next task is to translate the stream of operating costs into a pres-
ent value equivalent. The operating costs are assumed to be displaced ex-
penses for the firm and therefore do not require adjustment by the shadow
price of capital. These expenses represent displaced consumption and are dis-
counted at the social rate of time preference. This can be done with the
same formula as above:
6-22
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
1 -
P = A
0 * On| ,
where A is how a constant stream of annual operating costs. Table 6-7 shows
the present value of operating costs calculated with the same range of dis-
count rates used in the benefits calculation.
Table 6-7. Present Value of Operating Costs
Discount rate
(percent)
Present value of
operating costs
(million $)
2
4
6
10
37.1
29.7
24.3
17.2
Table 6-8 shows the present value of total project costs, which range
from 76 to 96 million 1981 dollars.
Table 6-8. Total Project Costs
Discount
rate
(percent)
2
4
6
10
Capital3
59.2
59.2
59.2
59.2
Operating
37.1
29.7
24.3
17.2
Total
(million $)
96.3
88.9
83.5
76.4
Private capital costs of the steel companies are adjusted
by shadow price of capital.
Note: Even if EPA assistance is received for plant construction, the society
still incurs roughly the same costs because these EPA funds would be di-
verted from their next best alternative use. There would perhaps be some
6-23
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
difference because Federal and local dollars might have different forgone op-
portunities, but the difference is probably minor for purposes of this assess-
ment.
table 6-9. Benefits and Costs of Attaining Fish and Wildlife
Propagation Use Designation
Cost/benefits
Quantity
Monetary value
(million $)
Benefits
Recreation
(fishing, swimming,
near water)
Enhanced ecological
diversity
User/nonuse—intrinsic
benefits to users and
nonusers
Costs
Capital .and operating
(Firms; City B)
Environmental
Not quantified
Not quantifiable
Not quantified
for this study
Not relevant
None
35 to 92
Not monetizable
Not monetized
for this study
76 to 96
0
VI. Assess the Plausibility of Total Benefits and Costs
To .assess the plausibility of benefits and costs, construct an array and
compare them. For example, Table 6-9 is an array showing both the benefits
and the costs of attaining the fish and wildlife use designation. The ranges
of monetized benefits and monetized costs overlap, nonmonetized benefits are
positive, and there are no nonmonetized costs. The upper end of the monet-
ized benefits is slightly less than the upper end of the range of monetized
costs. The lower end of the monetized benefits is considerably less than the
lower end of the cost range. Thus, the use designation decision would have
to consider whether the intrinsic benefits and nonquantified, enhanced ecolog-
ical diversity are greater than the difference between lower benefits and cost
bounds. When the suggested 2 percent social rate of time preference is
used, the monetized benefits are approximately equal to costs.
Table 6-10 shows expected benefits from the use designation change and
their distribution among area residents and visitors. In some instances, de-
scribing the distribution of benefits among more narrowly defined groups may
be desirable. For example, assessments involving a river segment that forms
the boundary between two States may require more precise distribution infor-
mation because the political issues are more complex when more than one State
is involved.
f 6-24
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Table 6-10. Distribution of Benefits and Costs
Benefit/cost
Distribution
Recreation
Enhanced ecological
diversity
Nonuse/nonuser
benefits
Navigation
Increased operating and
capital cost for steel
plants
Capital and operating
costs of advanced
treatment plant
Users of River 1--primarily recreators
from City A and City B
To some extent same as above, but
probably accrues to society as a whole
including future generations.
Residents of river basin. Visiting
recreators who have bequest motive
for future generations.
Shippers, boat owners, purchasers
of materials shipped along river.
Small magnitude in this study.
American and Riverton stockholders
to the extent the costs are not
passed on to customers. Given the
elasticity of demand for steel and
market shares of these firms, some
costs will be passed on to users of
steel products. U.S. taxpayers may
also bear some costs because of special
depreciation treatment for pollution
control expenditures.
$15 per user of water and sewer
service in City B. This may be
lower if construction grant Is obtained;
then U.S. taxpayers will share in
the burden.
To gain additional perspective on the distribution of benefits and costs,
the breakdown of recreational benefits by different income groups Is shown
below:
Income ($)
0 to 10,000
10,000 to 20,000
20,000 to 30,000
30,000 to 40,000
40,000+
Benefits (percent)
15
25
35
20
5
In this example, 75 percent of the benefits accrue to households with incomes
of less than $30,000/year. Decisionmakers may view this use change as more
6-25
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
desirable because of its larger share of benefits going to low and middle in-
come groups.
The expected costs of the programs also are arrayed in Table 6-10. The
basic economic principle underlying the distribution of the industrial operating
and capital cost is the responsiveness of demand for the industry's or firm's
products—the larger the responsiveness of quantity demanded to price
changes, the smaller will be the ability of the firm or industry to pass along
cost increases to its customers. The stockholders, workers, and other re-
source suppliers to the firm will have to bear the cost increases that are not
passed along through lower dividends or reduced wages. For the steer firms
in this example, the degree of sales responsiveness to price changes will be
influenced by their share of the market, the demands for products which use
the type of steel produced, competition from foreign sources of steel,-and
competition from substitutes (aluminum).
The municipal waste treatment costs likely will be borne by the residents
of City B. Residents will include current residents and future residents who
move, or are annexed, into the city's water or sewer service system. If the
city receives a Federal construction assistance grant, then the largest share
of the capital costs is distributed among a much larger group—the Federal
taxpayers.
•*
This example shows how distribution issues are integrated with the other
steps in the assessment process. They are described and listed but do not
affect the measurement of either benefits or costs. Finally, the exact nature
of the distribution descriptions will depend on the complexity of the issues In
the assessment itself.
6.5 SUMMARY
The strength of benefit-cost assessment is its ability to consist-
ently organize information for a wide range of applications.
Qualitative assessments, using primarily descriptive informa-
tion, are sufficient in many water quality applications because
of the sizes of potential benefits and costs.
Defining the water quality action to be evaluated in the
benefit-cost assessment determines the baseline and suggests
the level of complexity in the assessment.
Translating effects of a water quality action into beneficial
outcomes and costs requires an understanding of the linkages
between each element.
Valuing beneficial effects should be based on individual's will-
ingness to pay for them.
Valuing detrimental effects should be based on opportunity
cost to society of the effects.
6-26
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Comparing total benefits and total costs may involve qualita-
tive, quantitative, or monetized information.
Assessing the plausibilty of benefits and costs may involve test-
ing the sensitivity of assumptions made in estimating benefits
and costs, or in selecting a discount rate.
Describing the distribution of benefits and costs may provide
valuable data for the decisionmaker.
6-27
-------
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
CHAPTER 7
REFERENCES
Anderson, J. E., 1974, "A Note on Welfare Surpluses and Gains from Trade
in General Equilibrium," American Economic Review, Vol. 64, No. 4,
September 1974, pp. 758-62.
Ayres, R. V., and A. V. Kneese, 1969, "Production and Consumption Exter-
nalities," American Economic Review, 59, No. 3, June 1969, 282-97.
Bishop, R. C., and T. A. Heberlein, 1979, "Measuring Values of Extra-Market
Goods: Are Indirect Measures Biased?" American Journal of Agricultural
Economics, Vol. 6, December 1979, pp. 926-30.
Bockstael, Nancy E., and Kenneth E. McConnell, 1981, "Theory and Estimation
of the Household Production Function for Wildlife Recreation," Journal of
Environmental Economics and Management, Vol. 8, September 1981,
pp. 199-214.
Brookshire, David S., Mark A. Thayer, William D. Schulze, and Ralph C.
d'Arge, 1982, "Valuing Public Goods: A Comparison of Survey and
Hedonic Approaches," American Economic Review, Vol. 72, March 1982,
pp. 165-77.
Carlton, D. W., 1979, "Valuing Market Benefits and Costs in Related Output
and Input Markets," American Economic Review, Vol. 69, No. 4, Septem-
ber 1979, pp. 688-97.
Chemical Engineering, published by McGraw-Hill, New York
Cronin, Francis J., 1982, Valuing Njanmarket Goods Through Contingent Mar-
kets, prepared for U.S. Environmental Protection Agency, Pacific North~
west Laboratory, Richland, Washington, September 1982.
Davidson, P., F. G. Adams, and J. Seneca, 1966, "The Social Value of Water
Recreational Facilities Resulting From an Improvement in Water Quality:
The Delaware Estuary" in Allen V. Kneese and Stephen G. Smith, eds.,
Water Research, Johns Hopkins Press for Resources for the Future, Inc.,
1956T "
Desvousges, William H., V. Kerry Smith, and Matthew P. McGivney, 1983, A
Comparison of Alternative Approaches for Estimating Recreation and
Related Benefits of Water Quality Improvements, report prepared'for U.S.
Environmental Protection Agency, Research Triangle Institute, Research
Triangle Park, North Carolina, March 1983.
7-1
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Deyak, Timothy A., and V. Kerry Smith, 1978, "Congestion and Participation
in Outdoor Recreation: A Household Production Approach," Journal of
Environmental Ecgn£rnics and Management, Vol. 5, March 1978, pp. 63-80.
Dwyer, J, F., J. R. Kelly, and M. D. Bowes, 1977, Improved Procedures for
Valuation of the Contribution of Recreation to National Economic Develop-
ment, Urbana-Champaign: University of Illinois, 1977.
Engineering News- Record, published by McGraw-Hill, New York.
Fox, I. K., and O. C. Herfindahl, 1964, "Attainment of Efficiency in Satisfy-
ing Demands for Water Resources," American Economic Review, Vol. 54,
No. 2, May 1964, pp. 198-207.
"v
Freeman, A. Myrick, Ml, 1979, The Benefits of Environmental
Theory and Practice, Baltimore: Johns Hopkins Press for Resources for
the Future, Inc., 1979.
Guthrie, K. M., 1974, Process Plant Estimating, Evaluation, and Control ,
Solana Beach, California: Craftsman Book Company of America, 197T!
Just, Richard E., Darrell L. Hueth, and Andrew Schmitz, 1982, Applied Wel-
fare Economics arid Public Pqljcy., Englewood Cliffs, New Jersey: Prentice
Hall, 1982.
Larson, S., 1981, Economics .Analysis of Water Quality Improvementsi Missis-
sippi River, Minjiea|jgljs-St.^PauJ Twin Cities Area, Minnesota Pollution
Control Agency, Roseville, Minnesota7~T§8T.
Lind, R. C., K. J. Arrow, G. Corey, P. Dasgupta, A. Sen, t. Stauffer,
J. E. Stiglitz, J. Stockfish, and R. Wilson, 1982, Discounting for Time
and Risk in Energy Policy, Baltimore: Johns Hopkins, 1982.
R. S. Means Company, Inc., 1981, Building Construction Cost Data, Duxbury,
Massachusetts: R. S. Means Company, Inc. , 1981.
Michel, Robert L., "EPA Municipal Wastewater Treatment Plant Operation and
Maintenance Cost Index," U.S. Environmental Protection Agency,
Washington, D.C., issued periodically.
/"
Michel, Robert L., "EPA Sewage Treatment Plant and Sewer Construction Cost
Index, U.S. Environmental Protection Agency, Washington, D.C., issued
periodically.
Mitchell, R. C., and R. T. Carson, 1981, Ah Experiment in Determining Will-
ingness to Pay for National Water Quajjty I improvements , draft report
prepared for U.S. Environmental Protection Agency, Resources for the
Future, Inc., Washington, D.C., June 1981.
Municipal Finance Officers Association and Peat, Marwick, Mitchell and Co.,
1982, Financial Capability Guidebook, draft, prepared for U.S. Environ-
mental Protection Agency, Washington, D.C., May 1982.
7-2
-------
I
1
1
i
i
i
i
i
i
i
'—!
i
i
i
i
i
I
i
i
Rae, Douglas A., 1981a, Visibility Impairment at Mesa Verde National Park:
An Analysis of Benefits and Costs of Controlling Emissions jn_ the Four
Corners Area, prepared for the Electric Power Research Institute,
Charles River. Associates, Cambridge, Massachusetts, 1981.
Rae, Douglas A., 1981b, Benefits of Improving Visibility at Great Smoky
National Park, draft report prepared for Electric Power Research Insti-
tute, Charles River Associates, Cambridge, Massachusetts, December
1981.
Richardson Engineering Services, Inc., 1977, Process Plant Construction Esti-
mating Standards: The Richardson Rapid System. Solana Be€ch, Cali-
fornia: Richardson Engineering Services, Inc., 1977.
Rowe, R. D., and L. G. Chestnut, 1981, Visibility Benefits Assessment Guide-
book, prepared for U.S. Environmental Protection Agency, Abt West,
Denver, Colorado, March 1981. X^
Schulze, W. D., R. C. d'Arge, and D. S. Brookshire, 1981, "Valuing Envi-
ronmental Commodities: Some Recent Experiments," Land Economics,
Vol. 57, No. 2, May 1981, pp. 151-73.
Smith, V. Kerry, 1975, "Travel Cost Demand Models for Wilderness Recreation:
A Problem of Non-Nested Hypotheses," Land Economics, Vol. 51, May
1975, pp. 103-11.
U.S. Department of Labor, Bureau of Labor Statistics, 1982, Producer Prices
and Price Indexes Data for August 1982, Washington, D.C., 1982.
U.S. Environmental Protection Agency, 1976, Areawide Assessment Procedures
Manual, Vol. Ill, EPA-600/9-76-014, U.S. Environmental Protection
Agency, Cincinnati, Ohio, 1976.
U.S. Environmental Protection Agency, Office of Water Program Operations,
Office of Research and Development, 1980a, Innovative and Alternative
technology .Assessment Manual, LD-53, U.S. Environmental Protection
Agency, Washington, D.C., February 1980.
U.S. Environmental Protection Agency, Effluent Guidelines Division, 1980b,
Development Document for Effluent Limitations Guidelines and Standards
for the Pulp, Paper and the Builder's Paper and Board Mijjs^ Point
Source Categories, EPA 44Wl-80/025-b, U.S. Environmental Protection
Agency, Washington, D.C., December 1980.
U.S. Environmental Protection Agency, 1981, Process Design and Cost Estimat-
ing Algorithms for the Computer Assisted Procedure for Design and Eval-
uation of Wastewater Treatment Systems (CAPDET), U.S. Environmental
Protection Agency, Washington, D.C., January 1981.
Vaughan, W. J., and C. S. Russell, 1982, "Freshwater Recreational Fishing:
The National Benefits of Water Pollution Control,'1 Washington, D.C.:
Resources for the Future, Inc., November 19§2T
7-3
------- |