GEF
DISCLAIMER:
The contents of this book do not necessarily reflect the views and policies of UNDP/GEF.
Information presented was provided by the author. Use of appropriate
byline/photo/image/data credit is requested. UNDP/GEF YSLME Project provides no
warranty, expressed or implied, as to the accuracy, reliability or completeness of furnished
data. Some of the information in the reports may contain references to information created
and maintained by other organisations. Please note that UNDP/GEF YSLME Project does
not control and cannot guarantee the relevance, timeliness, or accuracy of these outside
materials.
For reference purposes, this report may be cited as:
UNDP/GEF 2008. Guideline for Economic Analyses of Environmental Management Actions
for the Yellow Sea. UNDP/GEF YSLME Project, Ansan, Republic of Korea (35 pages).
GEF
UNDP/GEF PROJECT ENTITLED "REDUCING ENVIRONMENTAL STRESS IN THE
YELLOW SEA LARGE MARINE ECOSYSTEM"
Guideline for Economic Analyses of Environmental Management
Actions for the Yellow Sea
Isao Endo
UNDP/GEF Project entitled "Reducing Environmental Stress in the Yellow Sea Large
Marine Ecosystem" (YSLME Project)
Project Management Office
Ansan, Republic of Korea
United Nations Development Programme/
Global Environment Facility
July 2008
Abstract
This guideline describes how to assess the economy of management actions to
conserve marine and coastal resources. The guideline discusses the basics of
environmental valuation, explaining economic value, negative externalities, and
valuation techniques. The methodology of cost-benefit analyses is then discussed
with the actions' benefits and costs defined and with the multiple-step analytical
procedure explained. The guideline focuses on the Yellow Sea ecosystem, although
most concepts and techniques that are discussed here may be applicable to other
marine and coastal ecosystems in various regions. The relevant information for this
guideline came from books, articles in periodicals, government documents, and
publications by international organisations.
iii
iv
Table of Contents
FOREWORD............................................................................................................. VII
ACKNOWLEDGEMENTS .......................................................................................... IX
1 INTRODUCTION..................................................................................................1
1.1 BACKGROUND ........................................................................................1
1.2 TOPICS ..................................................................................................1
1.3 TARGET AUDIENCE .................................................................................2
1.4 ORGANISATION.......................................................................................2
2
BASIC ENVIRONMENTAL VALUATION .............................................................5
2.1 ECONOMIC VALUE OF GOODS AND SERVICES ...........................................5
2.2 WELFARE LOSS DUE TO NEGATIVE EXTERNALITIES...................................5
2.3 VALUATION TECHNIQUES ........................................................................7
2.3.1
Empirical technique for market goods and services ...........................10
2.3.2
Techniques for non-market goods and services.................................12
2.3.2.1 Travel cost method (zonal travel cost method)...............................12
2.3.2.2 Contingent valuation method (dichotomous choice method)..........15
3
COST-BENEFIT ANALYSES OF ENVIRONMENTAL MANAGEMENT
ACTIONS ...........................................................................................................19
3.1 BASIC FRAMEWORK OF COST-BENEFIT ANALYSES ..................................19
3.1.1
Change in economic value due to environmental degradation...........19
3.1.2
Benefit of management actions as prevented loss in economic value
............................................................................................................21
3.1.3
Cost of management actions..............................................................21
3.1.4
Cost-benefit analyses for decision-making.........................................22
3.2 PROCEDURE OF COST-BENEFIT ANALYSES ............................................23
4 SUMMARY.........................................................................................................29
REFERENCES...........................................................................................................30
APPENDIXES ............................................................................................................31
APPENDIX 1: CONSUMER SURPLUS ESTIMATED BY MARKET DEMAND CURVE FOR THE
SITE .....................................................................................................31
APPENDIX 2: LOGIT MODEL.......................................................................................34
APPENDIX 3: INDIVIDUALS' CONSUMER SURPLUS ESTIMATED BY CVM ........................35
v
List of Figures
FIGURE 1: ECONOMIC VALUE OF GOODS AND SERVICES ..................................................5
FIGURE 2: DEADWEIGHT LOSS DUE TO NEGATIVE EXTERNALITIES.....................................7
FIGURE 3: FITTING LINEAR DEMAND AND SUPPLY CURVES TO DATA ................................11
FIGURE 4: ESTIMATED DEMAND CURVE FOR A HYPOTHETICAL RECREATIONAL SITE.........14
FIGURE 5: ESTIMATED RELATIONSHIP BETWEEN THE BID PRICES AND THE PROBABILITY
FOR INDIVIDUALS TO REPLY "YES" TO ACCEPT THE PRICES .............................17
FIGURE 6: SHIFT IN SUPPLY FOR COMMERCIAL FISH DUE TO THE DECLINE IN FISH STOCK 19
FIGURE 7: SHIFT IN DEMAND FOR A HYPOTHETICAL RECREATIONAL SITE DUE TO WATER
DEGRADATION ..............................................................................................21
FIGURE 8: COST-BENEFIT ANALYSIS OF ENVIRONMENTAL MANAGEMENT ACTIONS...........23
FIGURE 9: ESTIMATED MARKET DEMAND CURVE FOR A HYPOTHETICAL RECREATIONAL
SITE .............................................................................................................33
List of Tables
TABLE 1: TECHNIQUES FOR VALUING ENVIRONMENTAL GOODS .......................................7
TABLE 2: SUGGESTED TECHNIQUES FOR VALUING ENVIRONMENTAL GOODS ...................9
TABLE 3: DEMAND AND SUPPLY FOR COMMERCIAL FISH ...............................................10
TABLE 4: TRAVEL COST TO A HYPOTHETICAL RECREATIONAL SITE (A SAMPLE VISITOR)..13
TABLE 5: TRAVEL COST TO A HYPOTHETICAL RECREATIONAL SITE (FIVE SAMPLE
VISITORS) ....................................................................................................13
TABLE 6: CONSUMER SURPLUS FOR A HYPOTHETICAL RECREATIONAL SITE ...................14
TABLE 7: SAMPLED INDIVIDUALS' WILLINGNESS TO PAY FOR COASTAL SITE
REHABILITATION ...........................................................................................16
TABLE 8: DECLINE IN THE NUMBER OF VISITS TO A HYPOTHETICAL RECREATIONAL SITE
DUE TO ENVIRONMENTAL RESOURCE DEGRADATION ......................................20
TABLE 9: CATEGORIES OF EXPECTED BENEFITS AND COSTS OF MANAGEMENT ACTIONS TO
REDUCE HYPOTHETICAL RECLAIMED LAND AREA ............................................25
TABLE 10: BENEFITS OF MANAGEMENT ACTIONS FROM A HYPOTHETICAL CASE...............27
TABLE 11: SENSITIVITY-ANALYSIS RESULTS: NET PRESENT VALUE OF MANAGEMENT
ACTIONS FROM A HYPOTHETICAL CASE ..........................................................28
TABLE 12: MARKET DEMAND FOR A HYPOTHETICAL RECREATIONAL SITE ........................32
TABLE 13: ESTIMATED INDIVIDUALS' CONSUMER SURPLUS FOR COASTAL SITE
REHABILITATION ...........................................................................................35
vi
Foreword
This Guideline was prepared under the UNDP/GEF Project entitled "Reducing
Environmental Stress in the Yellow Sea Large Marine Ecosystem," known as the
YSLME Project.
The YSLME Project aims to promote the sustainable development of the Yellow Sea,
reducing human-induced stress on its ecosystem. To achieve this goal, the Project
takes an ecosystem-based approach, conducting a number of activities, including
scientific research, policy planning, capacity building, and awareness campaigns.
The Project facilitates environmentally-sustainable management and use of the
Yellow Sea by developing the Strategic Action Programme (SAP), a regional
environmental strategy with management targets and actions. Given government
endorsement, the SAP will contribute to not only conserving the Yellow Sea
ecosystem, but also enabling people to continue benefiting from the abundant gifts
and services that the Yellow Sea provides.
Economic analyses play an important role in the ecosystem-based management of
marine and coastal resources in the Yellow Sea. The analyses help in ensuring that
environmental policies and management actions are economically efficient and,
therefore, that those actions could attain expected results in a cost-effective manner.
The analyses will greatly contribute to improving ecosystem management. It is
expected that the economic analyses, in particular the cost and benefit analyses,
would become important management tools to evaluate the impacts of management
actions to be implemented in the marine and coastal areas of the Yellow Sea.
I believe that this Guideline will be useful for all those who deal with marine and
coastal development and management in the Yellow Sea as well as in other regions.
Yihang Jiang
Project Manager
UNDP/GEF YSLME Project
vii
Acknowledgements
I would like to thank many people who contributed to preparing this Guideline. I
thank experts of the YSLME Project's Regional Working Group for the Investment
Component, especially Dr. Sukjae Kwon and Prof. Jingmei Li, for providing useful
information about the economy of marine resources. Discussions with them enabled
me to shape the structure of this Guideline. I thank Dr. Francis Vorhies and Dr. Di
Jin for reviewing the draft and providing me with valuable advice. I thank Dr.
Dohoon Kim for providing his advice, explaining fishery economics in a simple way. I
thank Dr. Thuy Truong Dang for his instruction on valuation techniques. I thank Dr.
David Weimer for explaining the travel cost method in detail. Their contribution
greatly helped in refining the Guideline. I thank Dr. Xianshi Jin, Dr. Jian Guang Fang,
and Dr. Jihong Zhang for providing me with data and advice on fisheries and
mariculture in the Yellow Sea. I thank Ms. Sue Prasad for reviewing the draft.
Finally, I would like to thank people at the Project Management Office, especially Mr.
Yihang Jiang, Ms. Connie Chiang, and Dr. Mark Walton for their comments and
encouragement. They spanned the gap in my knowledge in the field of natural
science. All the support I obtained from these people and others made it possible to
prepare this document. However, any mistakes that remain are my own.
ix
Guideline for Economic Analyses of Environmental Management Actions for
the Yellow Sea
1 Introduction
1.1 Background
Marine and coastal ecosystems suffer from serious environmental degradation which
is attributable to various anthropogenic causes. The Yellow Sea ecosystem, a water
area adjacent to China and the Korean Peninsula, has experienced for a long time a
range of problems such as water quality degradation, declined fish stock, and
biodiversity loss (Yellow Sea Large Marine Ecosystem Project [YSLME], 2000). The
loss of opportunities for recreation and tourism is also a major concern (YSLME,
2005a, Annex IV, p. 9). Anthropogenic activities such as fishing, mariculture, and
tourism might cause these problems (YSLME, 2005b, Annex IV, p. 3). To mitigate
these environmental problems, the UNDP/GEF Project entitled "Reducing
Environmental Stress in the Yellow Sea Large Marine Ecosystem," known as the
YSLME Project, was launched in 2004.
Bordering three countries: the Democratic People's Republic of Korea (DPRK), the
People's Republic of China (China), and the Republic of Korea (ROK), the Yellow
Sea ecosystem is the semi-enclosed body of water with an area of about 400,000
km2. The floor of the Yellow Sea, submerged post-glacially, is unique geologically.
The seafloor has an average depth of 44 meters with the maximum depth of about
100 meters. The slope of the seafloor is gentle near the Chinese continent while the
slope is steep toward the Korean Peninsula. The Yellow Sea is connected to the
East China Sea in the south, forming a linked circulation system. With its high
primary productivity,1 the Yellow Sea ecosystem supports substantial populations of
fish, invertebrates, marine mammals, and seabirds. In addition, people in the coastal
countries have benefited for hundreds of years from those abundant gifts from the
Sea (YSLME, 2000).
The Project aims to develop a Transboundary Diagnostic Analysis (TDA) and a
Strategic Action Programme (SAP)--guides to assist in alleviating Yellow Sea's
environmental problems. Analysing historical data and trends in the region, the TDA
prioritises environmental problems which have a transboundary nature; then, it
identifies major causes of the problems. The SAP outlines management actions to
solve the priority problems. With the endorsement from the Project's participating
countries (i.e., China and ROK), the management actions will be implemented.
The SAP development process includes feasibility studies of the suggested
management actions. The actions are examined in terms of their technical,
economical, and political suitability and viability. Cost-benefit analyses are employed
as tools to assess the economic feasibility of the actions.
1.2 Topics
This Guideline provides practitioners of marine and coastal environmental
conservation with a set of instructions on how to conduct cost-benefit analyses on
management actions to mitigate ecosystem degradation. The Guideline presents the
basics of environmental economics, explaining valuation techniques and analytical
procedures. To compose the Guideline, a number of books and articles from the
1 Primary productivity is the amount of carbon fixed by photosynthesis. In the oceans, this is
mainly due to the growth of micro-algae or phytoplankton.
1
literature were reviewed, including: Boardman, Greenberg, Vining, and Weimer
(2006); Grigalunas, Opaluch, Diamantides, and Brown (1995); and Lipton, Wellman,
Sheifer, and Weiher (1995). Those texts constitute the foundation of the Guideline.
What makes this Guideline unique is its focused and detailed description. There are
a number of writings available for cost-benefit analyses of environmental
commodities, that introduce a variety of valuation methods and summarise earlier
research as case studies. However, those texts do not provide enough details for
those who have a limited knowledge of economics to conduct the analyses.
Practitioners need more detailed information on methodology: What steps should the
analyses take? What data should be collected specifically? How should those data
be analysed econometrically? How should analytical results be used for ecosystem
management? This Guideline is composed to meet such a need by providing the
step-by-step procedure of the analyses and by focusing on a few most important
valuation methods.
1.3 Target audience
This Guideline targets a wide range of audiences, including not only economic
researchers of marine and coastal environmental conservation, but also policy-
makers, development planners, and natural scientists. For practitioners, the
Guideline provides a handy guide to conduct cost-benefit analyses of environmental
management actions. For decision-makers, the Guideline offers an easy reference
to assess, interpret, and apply analytical results to marine and coastal management.
The Guideline focuses on the Yellow Sea ecosystem; however, most concepts and
techniques that are discussed in this Guideline may be applicable to other marine
and coastal ecosystems in different regions.
To understand the contents of the Guideline, it is useful, though not necessary, to
have a good understanding of basic applied microeconomics and statistical analysis.2
Computer skills of operating spreadsheet programmes are a minimum requirement
for researchers to prepare the economic analyses presented in this Guideline;
however, the skills are not required for those who use mainly the analytical results.
To fully understand and apply the presented methods and statistical techniques to
the evaluation of management actions, especially if they are complex, readers are
recommended to consult the literature cited in this Guideline.3
1.4 Organisation
The Guideline deals mainly with two topics: (i) environmental valuation and (ii) cost-
benefit analyses. Chapter 2 describes the basics of environmental valuation,
defining the "value" of environmental goods and services in terms of economy. The
concept of consumer and producer surpluses is introduced, which forms the
economic value. The concept of externalities is then introduced; the chapter explains
negative externalities as a cause of welfare loss for the society as a whole because
they reduce the economic value of concerned commodities. Finally, the chapter
presents a detailed explanation about valuation techniques, providing hypothetical
cases with numerical examples.
2 Pindyck and Rubinfeld (1995) concisely explain the basics of multiple regression analysis, a
means to fit economic relationship to observed data (pp. 659-667).
3 Ecosystem Valuation, a website designed for non-economists, provides the basics of
environmental valuation comprehensively (http://www.ecosystemvaluation.org/). The website
also provides a useful link to a number of relevant websites.
2
Chapter 3 presents the essentials of cost-benefit analyses, using the concept and
techniques discussed in Chapter 2. Benefits and costs are defined in the context of
assessing the economy of management actions. Providing simple decision criteria,
the chapter explains how to use the results of economic analyses for environmental
decision-making. An eight-step procedure of cost-benefit analyses is presented with
examples. The procedure includes important components of economic analyses,
such as the net present value calculation and the sensitivity analyses. The Guideline
explains the concept of discounting, suggesting a specific rate for its calculation, to
incorporate the time factor if benefits and costs accrue over time.
3
2 Basic environmental valuation
2.1 Economic value of goods and services
The economic value of goods and services is defined as the sum of consumer
surplus and producer surplus. (For convenience, hereinafter, the term "good[s]"
includes both "good[s]" and "service[s]".) The "consumer surplus is the difference
between what a consumer is willing to pay for a good and what the consumer
actually pays when buying it" (Pindyck & Rubinfeld, 1995, p. 113). The producer
surplus is "the difference between the cost of producing a commodity [good] and the
revenue received by selling the commodity [good]" (Grigalunas et al., 1995, p. 25).
Graphically, the consumer surplus is an area between the demand curve and the
market price for the good. Meanwhile, the producer surplus is an area above the
supply curve up to the market price for the good (Figure 1).
Price
Supply
Consumer
surplus
P0
Producer
Demand
surplus
Q0
Quantity
Source: Pindyck & Rubinfeld, 1995, p. 278
Figure 1: Economic value of goods and services
The downward demand curve is derived from consumer behaviour: Consumers are
willing to buy more goods as their price becomes lower. The upward supply curve is
derived from producer behaviour: Producers (e.g., firms) are willing to produce more
goods as their price becomes higher. The supply curve shows the information about
firms' production cost (i.e., marginal/incremental valuable cost).
The economic value is maximised if goods are provided at the price and quantity
when the demand curve and the supply curve for goods intersect; Figure 1 depicts
such a condition. When the economic value is maximised, a society is well-off; in
other words, social welfare is maximised, at least in terms of economy.
2.2 Welfare loss due to negative externalities
The economic value of goods or the social welfare is not maximised when negative
externalities exist. The negative externalities are defined as a condition such that
5
"the agent responsible must not take account of the effect that it has on the other
party" (Markandya, Perelet, Mason, & Taylor, 2001, p.94).
To understand the concept of the negative externalities, consider water pollution
caused by steel production. (This example is adapted from Pindyck and Rubinfeld
[1995, pp. 624-626].) Suppose that a company produces pollutants as it produces
steel, discharging pollutants through wastewater into a river without treating them.
As a result, fish die or disappear and so fishermen operating downstream suffer from
catching fewer fish. This hypothetical example shows that river pollution costs not
the steel company, but the fishermen. The fishermen pay "cost" by losing the income
from catching fish because the company does not shoulder the cost of treating
wastewater. That is the case of negative externalities: An action taken by one party
(the steel company) negatively impacts other party (the fishermen). Those
externalities, as mentioned below, should be incorporated or "internalised" so as not
to cost the other party (or society) by avoiding excess production of goods, and
therefore pollutants.
Figure 2 shows negative externalities, following the above example. The company
produces steel at Q0 when the supply curve S (that describes the company's
production cost) intersects with the demand curve D for steel. The supply curve S
does not reflect the cost of controlling the pollution. However, such a cost actually
exists: Recall the "cost" paid by the fishermen in the example. The supply curve S*
represents the actual cost of supplying steel (i.e., the cost of both producing steel
and treating pollution). From the perspective of a society, steel should be produced
at Q* when the supply curve S* intersects with the demand curve D; then, the
economic value for the society as a whole is maximised. Note that Q* is less than
Q0. That is, without considering the pollution treatment cost, the company produces
more than it should from the perspective of the society. When the company
continues to produce steel at Q0, a loss called "deadweight loss" arises which the
society has to bear. The area marked with diagonal lines in Figure 2 represents the
deadweight loss due to the negative externalities caused by the excess steel
production (i.e., the difference between Q0 and Q*). The economic value for the
society as a whole is lessened by the deadweight loss. The total economic value of
producing steel at Q0 when the company does not consider the cost of controlling the
pollution is the difference between the area marked by ABC and the deadweight loss.
The society would not suffer from this loss if the pollution cost were internalised, and
the company produced less steel in the amount of Q*.
6
Price
Deadweight
S* S
B
loss
C
A
D
Q* Q0
Quantity
Source: Pindyck & Rubinfeld, 1995, p. 625
Figure 2: Deadweight loss due to negative externalities
2.3 Valuation techniques
Various techniques are available to measure the economic value of environmental
goods. Table 1 summarises common techniques, classifying them into four
categories: direct observable, direct hypothetical, indirect observable, and indirect
hypothetical. All those techniques require collecting and analysing field data (i.e.,
primary information source). Meanwhile, there is an approach known as "benefit
transfer" which uses "existing valuation information for one good or service to
estimate the value of a similar good or service" (Abt Associates Inc. [AAI], 2005, p. 1-
1). Unlike other techniques, the benefit transfer uses the findings of other existing
studies (i.e., secondary information source); therefore, the benefit transfer requires
less costs and time than other techniques.4
Table 1: Techniques for valuing environmental goods
Methods Observed
behaviour
Hypothetical
Direct Market
price
Contingent valuation
Simulated markets
Indirect Travel
cost
Contingent ranking
Hedonic property values
Hedonic wage values
Avoidance expenditures
Source: Tietenberg, 2003, p. 39
4 This is part of the reason that the benefit transfer is widely practiced (AAI, 2005, p. 1-1).
However, as Pagiola, Ritter, and Bishop (2004) point out, this approach is extremely
controversial because it has often been used inappropriately (p. 22).
7
One can estimate the economic value of goods, using their demand and supply
information. An idea behind the value estimation is straightforward, although
implementing the idea may not be easy. To estimate the economic value, first, one
should estimate the demand and supply curves of concerned goods by using the
methods described below in this section; then, one can calculate the area of the
consumer and producer surpluses of consuming/producing the goods.
If the goods are traded in the market, one can use the goods' market prices and
trading volumes to estimate the demand and supply curves. If the goods are not
traded in the market, however, one should use either the market information of
relevant goods or the information collected by surveys about consumer preference
for the goods concerned. It should be noted that if a target is market goods, one
should consider both the demand and the supply for the goods. However, if a target
is non-market goods, one can consider only the demand for the goods because non-
market goods such as recreational opportunities (e.g., scenic views) and biodiversity
have "no producer, or the consumer is both the producer and consumer" (Lipton et
al., 1995, p. 42).
The following sections discuss methods and procedures to estimate the demand and
supply for goods according to their nature of being traded in the market or not. The
focus is on the most appropriate techniques in the context of the Yellow Sea: the
empirical technique (referred to often as the market price method or the productivity
change method), the travel cost method, and the contingent valuation method. 5
Table 2 summarises those suggested techniques and their applications as described
in detail below.
5 Other methods such as the hedonic property value method are not discussed in detail in this
Guideline due to their limitation in data availability in the Yellow Sea region, though the
methods are frequently used in other regions, especially North America and Europe. The
detailed discussion of the benefit transfer, using values or functions estimated by existing
studies, is also not provided in this Guideline for similar reasons.
8
chnique:
chnique:
chnique:
te
te
te
gression
y
via
Reference
Regression analysis
Regression analysis
Logistic re
analysis
interviews
·
Statistical
·
Statistical
·
Statistical
·
Surve
trip
ation of
m
o pay
statistics
ed with
s
information
government
i
tors
Necessary data
Market price and
trading volume of
target good
associat
to target site
Wage infor
vis
district
Number of visits per
person
Number of visitors
willingness t
·
·
Cost
·
·
Local
·
·
·
Individuals'
·
Population
diversity
d on the
nsumer
rplus
on
d
on
o pay
d
co
an
an
9
data
data
data
cally
cally
data
cally
et
Procedure
cluding bio
Collect empirical
data on goo
mark
statisti
surplus and
producer su
Collect data
tourists
statisti
aggregate consumer
surplus
willingness t
statisti
aggregate consumer
surplus
1.
2. Analyse
3. Calculate
1.
2. Analyse
3. Calculate
1. Collect
2. Analyse
3. Calculate
al goods, in
chnique
od)*
environment
Table 2: Suggested techniques for valuing environmental goods
Valuation te
Empirical technique
Zonal travel cost method
Contingent valuation
method (dichotomous
choice meth
i
de range of
(e.g.,
s)
iew
Target goods
Market goods (e.g.,
commercial fish)
Non-market goods
scenic v
Notes: *Applicable to a w
2.3.1 Empirical technique for market goods and services
A procedure to estimate the demand and supply for market goods such as commercial fish
consists of the following four steps:
(1) Collect empirical data on the market prices and trading volumes of concerned goods;
(2) Collect empirical data on the marginal variable costs of producing the goods;
(3) Analyse statistically the market data collected in Step 1 to estimate the demand
curve; and
(4) Analyse statistically the cost data collected in Step 2 to estimate the supply curve.
Regression analyses are commonly used to estimate the demand and supply curves. One
can obtain functional forms of the curves, regressing the data by ordinary least squares.
(For more details on regression, see Pindyck and Rubinfeld [1995, pp. 659-667].) Widely-
used spreadsheet programmes have a function to conduct regression analyses. To illustrate
how to estimate the demand and supply for market goods, consider coastal commercial
fisheries as an example. Suppose that market information is collected as shown in Table 3.
(This example is adapted from Lipton et al. [1995, pp. 33-40].)
Table 3: Demand and supply for commercial fish
Price/Cost (USD per kg)
Demand (kg per day)
Supply (kg per day)
1 21,300 0
2 16,000
3,200
3 10,600
6,400
4 5,300
9,600
5 0
12,800
The price in USD and the demand in catch rate per day are those which generally prevail in
the market (i.e., the price and quantity that prevail "on average" or when market conditions
are "normal"). The supply is a quantity that is produced corresponding to the industry's
marginal variable cost that results from producing one extra unit of goods. In this example,
the marginal variable cost is the incremental cost to supply fish by one additional kilogram.
(See Pindyck and Rubinfeld [1995, pp. 42 and 198].)
Regression analyses provide the estimated demand and supply functions as follows. (For
simplicity, linear regression analyses are used.)
Demand : P = 5 - .
0
Q
000188
Supply : P = 1+
Q
000313
.
0
10
P and Q represent price and quantity, respectively.6 The t-statistics of the coefficients for the
quantity in the demand and supply functions are more than 1.96 in absolute value (-533 and
65535, respectively). That is, there is an association with 95 percent confidence between
the fish price and the quantity in demand for fish and between the marginal variable cost of
fishing and the quantity in supply for fish. The reason that the significant level of those
coefficients is high in this example is simply that the demand and supply data are prepared
purposely in such a way that there is a strong (linear) correlation between the price and
quantity. Figure 3 shows the estimated demand and supply curves that fit the data. (In
reality, data would not all lie exactly on estimated lines.)
6
5
B
S
C
(8000, 3.5)
4
g)
/k
E
D
S 3
U
e
(
r
ic 2
P
1
A
D
0
0
5000
10000
15000
20000
25000
Quantity (kg/day)
Source: Lipton et al., 1995, p. 38
Figure 3: Fitting linear demand and supply curves to data
According to the solution of the simultaneous equations of the demand and supply, the
intersecting point, C, is where the price is USD 3.5 per kg and the trading volume is 8,000 kg
per day. Given that, one can geometrically calculate the economic value as follows.
Economic value of commercial fisheries
= Area ABC
= Consumer surplus (Area EBC) + Producer surplus (Area AEC)
= (5 - 3 5
. )× 000
,
8
×1 2 + (3 5
. - )
1 × 000
,
8
×1 2
6 It is common practice for this kind of economic analysis to check with t-statistics whether estimated
coefficients are statistically significantly different from zero. As a rule of thumb, a coefficient is
different from zero if its t-statistic exceeds 1.96 in absolute value; then, one can claim that there is an
association with 95 percent confidence between a response variable and an explanatory variable(s).
Conventionally, t-statistics are presented with an estimated function to indicate the significant level of
estimated coefficients. Even if the estimated value of coefficients is not significantly different from
zero at the 95-percent confidence level, the value should be used for the purpose of cost-benefit
analyses because those coefficients may be the best estimate of the true value with given samples.
For more details on the statistical significance of estimated coefficients, see Boardman et al. (2006, pp.
328-329) and Pindyck and Rubinfeld (1995, pp. 662-663).
11
= USD 16,000 per day
Suppose that the total number of fishing days is 100 days a year; then, the economic value
of the commercial fish is USD 1.6 million per year (USD 16,000 x 100 days).
2.3.2 Techniques for non-market goods and services
If there is no available market information (i.e., price and trading volume) of target goods,
one should use either the information of other relevant market goods or surveyed information
about consumer preference for the target goods. In economics, it is common to call the
former way of using relevant good data as "revealed preference methods" and the latter way
of using survey data as "stated preference methods" (Freeman, 2003, p. 24). This section
discusses the travel cost method, a commonly-used revealed preference method; then, the
section describes the contingent valuation method, a commonly-used stated preference
method.
2.3.2.1 Travel cost method (zonal travel cost method)
The travel cost method (TCM) uses the cost information on how much people spend to
consume environmental goods as a proxy variable for their economic value. The method is
often applied to measure recreational services that environmental goods provide, such as
scenic views. The section below introduces the TCM, particularly the zonal TCM which uses
surveyed data of actual visitors with their departure points recorded and divided into areas or
"zones." The zonal TCM consists of three steps:
(1) Collect data on the travel cost information of visitors to a site (i.e., the travel cost of a
sample of visitors);
(2) Analyse the collected data statistically to estimate an individual visitor's demand
curve; and
(3) Calculate and aggregate the consumer surplus for visitors from different zones (i.e.,
extrapolate the consumer surplus for the sample to the entire population of the
visitors).
First, to reveal the environmental value of a recreational site, such as a beach, one should
collect the following information about visitors to the site (this example is adapted from
Boardman et al. [2006, pp. 354-361]):
· Travel
distance;
· Travel
time;
· Operating cost of vehicles (e.g., gasoline cost);
· Opportunity cost of the travel time (e.g., forgone time wage);
· Admission fee of the recreational site, if any;
(The above information gives the average total cost per person per visit.)
· Average number of visits per person per year; and
· Average number of visitors per year.
Suppose that a visitor who lives 2 km away from a beach (the target site to value) spends
half an hour each way to get to the beach (e.g., driving to the site, parking her car, and
walking to the entrance). She drives her car which consumes 15 cents per km of gasoline.
She pays USD 10 for the entrance fee to the site. Her hourly wage is USD 9.4; she would
get the salary of that amount if she uses her travelling time for work. She visits the beach 15
12
times per year. Then, the total travel cost of the visitor would be USD 20 per trip, as
calculated in Table 4.
Table 4: Travel cost to a hypothetical recreational site (a sample visitor)
Cost
(USD)
Reference
Opportunity cost
9.4
USD 9.4 x 0.5 hour x 2 trips
Operating cost
0.6
USD 0.15 x 2 km x 2 trips
Admission fee
10
One-time fee per trip
Total travel cost
20
Visits 15 times per year
Suppose that the information of four other visitors is also collected as shown in Table 5.
Each visitor is categorised by zone according to distance to the beach. In practice, it is
common to use local government jurisdictions as zones. The (average) total cost per person
is calculated in a similar way as described in Table 4.
Table 5: Travel cost to a hypothetical recreational site (five sample visitors)
Zone Travel
time
Travel distance
Average total
Average
(hours)
(km)
cost per person number of visits
per visit (USD)
per person per
year
A 0.5
2
20
15
B 1.0
30
30
13
C 2.0
90
65
6
D 3.0
140
80
3
E 3.5
150
90
1
Source: Boardman et al., 2006, p. 356
Second, regressing the data on the average total cost per person and the average number
of visits per person reveals the (representative) individual's demand curve for visits to the
beach as follows.
TC = 95 - V
5
where TC and V represent the travel cost per visit and the visits per person, respectively.
Figure 4 shows the estimated demand curve. (For simplicity, the above data were prepared
so that they would all lie exactly on the estimated line.)
13
100
90
E
Consumer surplus
)
of people from
D
80
D
S
Zone C
U
70
C,
C
T
Demand curve
60
TC = 95 5V
s
i
t
(
50
e
r vi
40
p
st
B
o
30
c
A
t
al
20
o
T
10
0
0
5
10
15
20
Visits per person (V)
Source: Boardman et al., 2006, p. 357
Figure 4: Estimated demand curve for a hypothetical recreational site
Third, using this figure, one can geometrically calculate consumer surplus for people from
different zones as Table 6 shows; for example, the consumer surplus for those who are from
Zone C is USD 90 per person ([USD 95 - USD 65] x 6 visits / 2). (See Column 2 in Table 6.)
Table 6: Consumer surplus for a hypothetical recreational site
Zone Average
number
Consumer
Number of
Consumer
of visits per
surplus per
visitors per year
surplus per
person per year person per year
Zone per year
(USD thousand)
(1)
(2)
(3)
(4) = (2) x (3)
A 15
562.5
10,000
5,625
B 13
422.5
10,000
4,225
C 6
90.0
20,000
1,800
D 3
22.5
10,000
225
E 1
2.5
10,000
25
Total
11,900
Source: Adapted from Boardman et al., 2006, p. 356
If population statistics are provided (i.e., the number of visitors), one can estimate consumer
surplus in each zone by multiplying the consumer surplus per person in each zone by its
corresponding population (for example, the consumer surplus of Zone C is USD 1.8 million
[USD 90 x 20,000 people]). (See Column 4.) Then, an analyst can estimate the total
consumer surplus for the visitors by summing those products: The total consumer surplus in
this example is USD 11.9 million per year. It is possible to estimate the total consumer
surplus by deriving the market demand curve for the site. For more information about that,
see Appendix 1 in this document.
14
2.3.2.2 Contingent valuation method (dichotomous choice method)
The contingent valuation method (CVM) estimates the economic value of environmental
goods by using survey results on individuals' willingness to pay (WTP) for the goods.
Providing plausible hypothetical scenarios (e.g., carefully describing the current and future
status of concerned ecosystems with and without conservation efforts), this method asks
respondents how much they would pay or whether they would pay a certain amount of
money to prevent environmental degradation. The CVM is applicable to a wide range of
environmental goods, including the goods that people have not yet used and/or will not use
(e.g., biodiversity) (Mitchell & Carson, 1989, p. 90).
According to Boardman et al. (2006), the CVM consists mainly of two groups of sub-
methods: the direct elicitation (nonreferendum) method and the dichotomous choice
(referendum) method (pp. 370-374). The former method includes the open-ended
willingness-to-pay method, the closed-ended iterating bidding method, and the contingent
ranking method. Those methods, at one time commonly used, are no longer in use due to
various limitations. The latter method was recommended as the method of choice in most
circumstances by a blue-ribbon panel of social scientists, that was convened by the National
Oceanic and Atmospheric Administration (Boardman et al., 2006, p. 370). The section
below, adapted mainly from Boardman et al. (2006) and Loomis (1988), illustrates how to
use the dichotomous choice method to measure the economic value of environmental
goods.
Suppose that a coastal site faces serious environmental problems. A local government that
has jurisdiction over the site decides to develop rehabilitation plans. The government also
decides to implement a study to understand the environmental value of the site, expecting
that the study results will contribute to developing the plans. To measure the value of the
site, one can employ the dichotomous choice method as follows:
(1) Collect data on individuals' WTP (i.e., the WTP of a sample of respondents from the
population) for environmental goods (in the example, the coastal site);
(2) Analyse the collected data statistically to estimate the individuals' WTP; and
(3) Calculate and aggregate the WTP to reveal the consumer surplus of having the
goods for the society as a whole (i.e., extrapolate the WTP for the sample to the
entire population).
First, one should collect data on individuals' (e.g., city residents and visitors who use the
site) WTP for rehabilitating the site. Using a questionnaire, interviewers can ask
respondents whether they would pay a certain amount of money to prevent environmental
degradation. Given one randomly drawn price, referred to as a "bid price," each respondent
is asked to state whether he would be willing to pay the price (Boardman et al., 2006, pp.
371-372). The following is a simplified sample question:
The site you are visiting is deteriorating due to lack of management and
maintenance. [Here, interviewers provide the detailed information about the site and
the environmental problems it faces.] Let us assume that the local government is
planning to rehabilitate the area and that, due to budget constraints, it is also
considering asking visitors to contribute to investment costs by paying an entrance
fee for a day visit. [Here, interviewers provide the detailed information about not only
the rehabilitation plans but also the consequences of implementing or not
implementing them.] Would you be willing to pay the following fee? [Here,
15
interviewers offer the respondent one bid price.] (Markandya, Harou, Bellu, & Cistulli,
2002, p. 453)
The data from the example survey are shown in Table 7. In this example, there are 12
respondents who are suggested different prices ranging from USD 5 to USD 60. If a
respondent replies "yes," that is recorded as 1. If he replies "no," that is recorded as 0
(Loomis, 1988, pp. 209-213).
Table 7: Sampled individuals' willingness to pay for coastal site rehabilitation
Bid price (USD per visit)
Response (1 = "yes," 0 = "no")
5 1
6 1
7 1
9 1
10 1
11 0
25 1
30 0
35 0
50 0
55 0
60 0
Source: Loomis, 1988, p. 210
Second, one should analyse the data statistically to estimate the individuals' WTP for the
site. The logistic regression, using the logit model, helps in estimating the relationship
between bid prices and responses, although there may be a number of other possible
models applicable. For more information about the logit model, see Appendix 2 in this
Guideline.
Using the logit model with the raw data in Table 7, one can estimate the individuals' WTP
function as follows (Loomis, 1988, p. 211).
P
RY = ln ( yes = .3321- 156
.
0
1
- yes )
BP
P
RY is the log of the odds ratio or the ratio of the probability that respondents would reply
"yes" to given bid prices, BP, to the probability that respondents would reply "no." To
estimate this equation, a statistical package is necessary. Taking the exponential of RY
gives:
exp( 321
.
3
- 156
.
0
BP)
P =
yes
1+ exp( .
3 321- .
0 156BP)
This estimated function explains the relationship between the bid prices and the probability
for individuals to reply "yes" to pay for rehabilitating the coastal site. For example, when the
bid price is 11 (i.e., BP = 11), the probability of individuals agreeing to pay that amount is
16
approximately 0.83 (Pyes = exp(3.321 0.156 x 11) / (1 + exp[3.321 0.156 x 11]) = 0.832).
Figure 5 shows the estimated logistic regression based on the data.
1.00
0.90
7,
( 0
.90
)
11
( ,
0.
8
3
)
0.80
y
e
s
)
P 0.70
y
e
s
"
(
0.60
i
ng " 0.50
e
ply
r 0.40
25
( ,
0.
3
6
)
y
of
0.30
ilit
b
0.20
r
oba
P 0.10
35
( ,
0.
1
1
)
0.00
0
10
20
30
40
50
60
Bid price (BP, USD)
Source: Adapted from Loomis, 1988, p. 212
Figure 5: Estimated relationship between the bid prices and the probability for
individuals to reply "yes" to accept the prices
Third, considering the estimated logistic regression function as the demand curve for the
coastal site concerned, one can estimate consumer surplus for the site. The area under the
function approximates the individuals' mean maximum WTP or the individuals' consumer
surplus for the site (Loomis, 1988, p. 212). According to Boardman et al. (2006), the area
can be calculated by the following five procedures:
First, divide the range of X [BP in the example] into equal segments of width n.
Second, calculate the probability of acceptance at each of these points. Third, find
the average acceptance value for adjacent pairs of points. Fourth, multiply each of
these averages by n. Fifth, sum all these products to get the estimate of the area
(pp. 397-398).
With the above procedures followed, the estimated individuals' consumer surplus for the site
is approximately USD 21.5. See Appendix 3 for more information on how to calculate the
individuals' consumer surplus. Then, one can estimate the aggregate consumer surplus or
the economic value of the site for the society as a whole by multiplying the individuals'
consumer surplus by the number of relevant individuals or households (Grigalunas et al.,
1995, p. 88; Lipton et al., 1995, p. 54). Assuming that there are 300,000 people concerned
17
in the example, one would estimate the economic value of the site at approximately USD 6.5
million per year (USD 21.5 x 300,000 people = USD 6,450 thousand).
18
3 Cost-benefit analyses of environmental management actions
3.1 Basic framework of cost-benefit analyses
3.1.1 Change in economic value due to environmental degradation
The economic value of environmental goods decreases because of environmental resource
degradation. For example, consider the decline in landings of commercial fish due to the
decline in fish stock, which is attributable to the overexploitation of the fish. The size of fish
catch depends on both the size of fish stock and the amount of fishing efforts (Tietenberg,
2003, p. 310). If the fish stock declines, fishermen have to increase fishing efforts (e.g.,
employ better equipment or more people) to maintain fish catch at the same level as before:
That costs fishermen. Put simply, reduced stock size increases fishing cost. As a result, the
supply curve of catching fish shifts to the left (Lipton et al., 1995, p. 37); one can recall that
the supply curve of producing goods is modelled as a function of a producer's marginal
variable cost (see Section 2.1). Figure 6, using the example discussed in Section 2.3.1 in
this Guideline, illustrates the shift in supply for commercial fish due to the decline in fish
stock.
6
Sless
5
B
S
E
(6723, 3.7)
4
g)
/k
D
C
S 3
U
e
(
r
ic 2
P
1
A
D
0
0
5000
10000
15000
20000
25000
Quantity (kg/day)
Figure 6: Shift in supply for commercial fish due to the decline in fish stock
Sless represents the supply for commercial fish when less stock is available due to
overexploitation, assuming that the cost of catching fish increases by 30 percent as an
example. The estimated function of the new supply curve, Sless, is as follows.
Supply
: P = 1+ .
0
Q
000407
less
Note that the coefficient for the quantity in demand in this new supply function with less stock
is 30 percent more than that in the original supply function with more stock (0.000407 =
0.000313 x 1.3). The demand and supply curves intersect at E where the price is USD 3.7
per kg and the trading volume is 6,723 kg per day. (Solving the simultaneous equations of
19
the two functions--the demand function [D] and the new supply function [Sless]--gives the
intersecting point. For the demand function, see Section 2.3.1.)
Given the above information, one can calculate the reduced economic value by taking the
difference between the economic values of goods before and after environmental resource
degradation. In the example, the economic value of commercial fisheries before
environmental degradation is USD 1.6 million per year (see Section 2.3.1). Meanwhile, the
economic value of commercial fisheries after environmental degradation is approximately
USD 13 thousand per day as calculated below, or USD 1.3 million per year on the
assumption that the total number of fishing days remains the same at 100 days a year (USD
13,446 x 100 days).
Economic value of commercial fisheries with less fish stock
= Area ABE
= (5 - )
1 × 723
,
6
×1 2
= USD 13,446 per day (Area AEC)
The reduced economic value of commercial fisheries is about USD 300 thousand per year;
that is the difference between USD 1.6 million and USD 1.3 million.
Environmental resource degradation also reduces the economic value of goods by affecting
the demand for them; for example, people might decide not to visit a beach where the water
is polluted. Suppose that the number of tourists to the beach in the example in Section
2.3.2.1 decreases by 10 percent as water quality degrades. Table 8 illustrates that change
as the 10-percent decline in the number of visits per person per year. For example, the
average number of visits per person from Zone B decreases by 10 percent from 13 times to
11.7 times.
Table 8: Decline in the number of visits to a hypothetical recreational site due to
environmental resource degradation
Zone Average Average
Average
Consumer
Number of
Consumer
total
number of
number of
surplus per
visitors
surplus per
cost per
visits per
visits per
person per
per year
Zone per
person
person per
person per
year (after
year (after
per visit year (before
year (after
degradation)
degradation)
(USD)
degradation) degradation)*
(USD
thousand)
A 20 15 13.5
506.3
10,000 5,063
B 30 13 11.7
380.3
10,000 3,803
C 65
6
5.4
81.0
20,000 1,620
D 80
3
2.7
20.3
10,000 203
E 90
1
0.9
2.3
10,000 23
Total 10,710
Notes: *10-percent decline in the number of visits assumed
Figure 7 shows the shift in demand, due to water degradation, for recreational opportunities
that the beach provides. D represents the original demand for the site, TC = 95 5V;
whereas, Dlow represents the reduced demand for the site due to low water quality, TC = 95
5.56V, estimated by ordinary least squares regressing the reduced number of visits on the
20
total cost per visit (the t-statistics of the coefficients of this estimated function are more than
1.96 in absolute value).
100
90
D)
80
US
70
C,
T
60
t
(
i
si
50
e
r v
40
D: TC = 95 -5V
p
st
o
30
c
t
al
20
o
T
Dlow: TC = 95- 5.56V
10
0
0
5
10
15
20
Visits per person (V)
Figure 7: Shift in demand for a hypothetical recreational site due to water degradation
One can calculate the annual consumer surplus per zone in the same way as described in
Section 2.3.2.1. For example, the annual consumer surplus for those who are from Zone A
is approximately USD 5 million ([USD 95 - USD 20] x 13.5 visits / 2 x 10,000 people = USD
5,063 thousand). The total consumer surplus for the visitors with the reduced demand is
USD 10.7 million per year, that is the sum of all the consumer surplus per zone. Then, the
reduced economic value of the beach is about USD 1.2 million per year with the difference
taken between the economic value under the original demand, USD 11.9 million, and that
under the reduced demand, USD 10.7 million.
3.1.2 Benefit of management actions as prevented loss in economic value
The benefit of management actions to mitigate environmental problems can be defined as
the prevented future loss measured in economic value. Recall in the example that the
reduced economic value of the commercial fisheries is about USD 300 thousand per year.
(See Section 3.1.1.) Suppose that a management action will be taken to prevent the decline
in fish stock by controlling overexploitation of the fish (e.g., reducing illegal fishing,
seasonal/area fishing ban) and that the action will reduce fishing cost so that the supply
curve of catching fish will shift to the right. For simplicity, assume in Figure 6 that the supply
curve shifts from Sless to S; then, the benefit of controlling overexploitation is USD 300
thousand per year; that is the prevented future loss in commercial fisheries.
3.1.3 Cost of management actions
The cost of management actions is relatively straightforward; it is defined as the cost
incurred to implement proposed actions. The cost consists of "both the direct costs of
implementing conservation measures, and the opportunity costs of foregone uses" (Pagiola
et al., 2004, p. 7). Direct costs may be divided into the following two categories: (i) the cost
to establish and initiate proposed management actions (installation cost); and (ii) the cost to
21
operate and maintain the actions (O&M cost). The opportunity costs are forgone future
benefits, which otherwise would be realised through other usages, due to the implementation
of the actions. For example, the opportunity cost of preserving mangrove forests is the
forgone profit from deforesting and converting the land for commercial use (Markandya et
al., 2001, p. 144). If one protected mangrove forests, he would give up future revenues from
the sale of agricultural crops, for instance, that were cultivated in the deforested area. In the
example of controlling the overexploitation of the fish (Section 3.1.2), the cost of
management actions may include the following: the direct costs of establishing and enforcing
laws and regulations, that include monitoring costs, and the opportunity cost of a fishing ban.
3.1.4 Cost-benefit analyses for decision-making
Analysing the benefits and costs of proposed management actions helps decision-makers
decide whether to implement the actions. Comparing the net benefits (i.e., the difference
between [gross] benefits and costs) of management actions under two scenarios, with or
without the actions, cost-benefit analyses address a research question: "What would happen
if conservation measures [management actions] were implemented [compared] to what
would have happened if they were not" (Pagiola et al., 2004, p. 19). The analyses then use
simple yet effective decision criteria: Comparing the gains (benefits) with the losses (costs)
of an action, if the former exceeds the latter, support the action; otherwise, oppose it
(Tietenberg, 2003, p. 19). With analytical results given, it is logical for decision-makers to
accept the proposed actions if the net benefits are positive, or to decline the actions if the net
benefits are negative.
Figure 8 illustrates the concept of a benefit-cost analysis under with or without scenarios.
Properly measured, the economic value of goods today may be illustrated as the leftmost
column in the figure. Suppose that the value will decrease in the future because of
environmental degradation; then, the value would be as shown in the next column to the
right. This situation with decreased value is a "baseline," which is defined as the "reality in
the absence of the regulation [management actions]" (U.S. Environmental Protection Agency
[U.S. EPA], 2000, p. 21). The difference in the amount of the economic value between today
and the future is the scale of predicted degradation. With management actions
implemented, however, this degradation might be less (third column from the left).
Comparing the results of the two scenarios, with or without management actions, would
reveal the benefit of the actions. In the subsequent cost-benefit analysis (the rightmost
column), the benefit of implementing the management actions is compared with the cost of
implementing them. The cost might consist of both direct costs and opportunity costs. If the
benefits exceed the costs, it is reasonable to support the management actions.
22
Predicted
degradation
Benefit
in future
of action
Econ.
value
Econ.
value
Econ.
value
Increase
in value
Economic
W ithout
Cost of
Cost of
value
action
action
action
today
in future
W ith
Cost-benefit
action
analysis
in future
Source: Adapted from Pagiola et al., 2004, pp. 13-21
Figure 8: Cost-benefit analysis of environmental management actions
It is important to note that the cost-benefit analyses should compare the benefit and cost
"with and without" the management actions, rather than "before and after" implementing
them. In other words, the analyses do not compare the economic value today and that in the
future with the actions. The reason for this is that many other factors may have changed in
the period of intervention (i.e., between today and sometime in the future); it is difficult to see
whether the increase in the economic value is attributable to the concerned management
actions or to other unaccounted factors (Pagiola et al., 2004, p. 19).
3.2 Procedure of cost-benefit analyses
The procedure of a cost-benefit analysis consists of the following eight steps (adapted from
Boardman et al. [2006, pp. 7-17]):
(1) Specify management actions to analyse;
(2) Predict future environmental degradation;
(3) List expected benefits and costs of the actions;
(4) Predict the benefits and costs quantitatively;
(5) Monetise the benefits and costs;
(6) Calculate the net present value of the benefits and costs;
(7) Conduct sensitivity analyses; and
(8) Make recommendations.
To explain each step specifically, image a hypothetical case as follows. There is a coastal
development plan to convert a wetland into various industrial usages. The development is
expected to bring economic profits to a local community. However, there is a concern about
23
the adverse impact of the development on the ecosystem in the proposed development site
and on the local economy near the site, such as coastal fisheries and tourism. The site
provides habitat for unique marine wildlife, including those in danger of extinction. The
wildlife would disappear if the plan were materialised. Additionally, the development might
pollute the seawater and cause a decline not only in coastal fish stock and catch, but also in
beach bathing areas/opportunity. Considering the above situation, the local government
decided to take management actions both to reduce the converted wetland area and to
control pollutants from the industries on the reclaimed land. The government also decided to
conduct a cost-benefit analysis of those actions to see whether they would be justifiable
economically. Using the above hypothetical case, the following sections explain the eight
steps for the cost-benefit analysis.
Step 1: Specify management actions to analyse
First, one should specify a set of management actions to analyse. In the hypothetical
example, the management actions are to reduce the reclaimed land area and the pollution.
As mentioned above in this chapter, cost-benefit analyses compare the net benefits of taking
management actions (with scenario) to that of taking no action (without scenario).
Step 2: Predict future environmental degradation
Second, one should predict likely environmental degradation in the future if no action is
taken. An estimated environmental value of goods with the predicted future loss is then
considered as a baseline to be compared with an estimated increased environmental value
of goods as a result of management actions. The prediction might require scientific
knowledge (e.g., environmental modelling).7
Step 3: List expected benefits and costs of the actions
Third, one should identify expected benefits from and costs of taking proposed actions. The
benefits of the actions are the difference between the economic value of goods under a
without-action scenario (baseline) and that under a with-action scenario. The costs of the
actions are all expenses incurred to install, operate, and maintain the actions. Those costs
might include opportunity costs caused by taking the actions.
In this example, the anticipated benefits of reducing the reclaimed land area and the
pollution may be an increase in the number of marine wildlife, coastal fish stock, and beach
tourists. Meanwhile, the anticipated costs may include not only the direct costs of
administering regulations to reduce the reclaimed land area (e.g., compliance monitoring
and enforcing the regulations) and of installing, operating, and maintaining pollution control
devices, but also the opportunity cost of forgone future benefits that would be realised if the
reclaimed area were not reduced. Table 9 summarises the benefits and costs expected as a
result of taking the actions.
7 Bioeconomic modelling might be useful to assess the economy of management actions such as
fisheries management. Having developed a bioeconomic model for red grouper fishery, Kim (2003)
evaluated the effect of management actions to recover fish stock, including a total allowable catch and
a five-month closure period.
24
Table 9: Categories of expected benefits and costs of management actions to reduce
hypothetical reclaimed land area
Benefit Cost
Increase in the number of:
Direct cost:
· marine wildlife
· regulation cost (e.g., compliance
· coastal fish stock
monitoring and enforcing cost)
· beach tourists
· installation, operation, and
maintenance cost of pollution
controlling facilities
Opportunity cost:
· forgone future benefits if the reclaimed
land area be not reduced
Step 4: Predict the benefits and costs quantitatively
Fourth, one should quantitatively predict at this stage the benefits and costs of management
actions in terms of their magnitude, not monetary value. On one hand, as was the case in
Step 2, predicting the benefits may require environmental modelling as well as socio-
economic survey to reveal cause-and-effect relationships between the actions (cause) and
the benefits of them (effect). On the other hand, to estimate the costs, there are three
approaches: survey approach, engineering approach, and combined approach with the
above two approaches (Tietenberg, 2003, pp. 47-48). The survey approach is to ask those
who know the most about the proposed management actions; the engineering approach is to
use general engineering information. The combined approach collects information on
possible technologies as well as on special circumstances; then, it derives the actual costs
of those technologies with the special circumstances considered. The combined approach is
preferable because it provides balanced information while minimising the problems of the
other two approaches.
In the example, an analyst should estimate the benefits by predicting how much marine
wildlife, coastal fish stock, and beach tourists would increase as a result of reducing the
reclamation area and pollution. Environmental modelling would help in estimating those
increases by predicting the relationship not only between the wetland area as habitats and
the marine animals, but between the pollution caused by the industry located on the
reclaimed land and the fish stock. Socio-economic survey is necessary to reveal the
relationship between the pollution and the number of tourists, predicting how many tourists
would visit the beach if the pollution were to decrease. The cost estimation in the example
requires interviews with those who know the most about administering the regulations and
developing the reclaimed land for industrial use. It is also necessary to evaluate specific
pollution control technologies by collecting information on possible technologies as well as
on special circumstances facing firms or areas where the technologies are introduced. The
information source may include the following: local government agencies which deal with
coastal management and development, land developers, manufacturers of pollution control
devices, operators of existing pollution control facilities, technical people of local coastal
industries, and universities with expertise in relevant fields.
25
Step 5: Monetise the benefits and costs
Fifth, one should place monetary values on the benefits and costs of management actions,
using techniques described in this Guideline. To measure the benefits, there are three
valuation techniques suggested in Section 2.3: empirical technique, zonal TCM, and CVM.
Using those techniques, one can estimate the economic values of goods without
management actions, or the baseline. Given the information obtained from Step 4 about the
benefits of management actions in "impacts," then, an analyst can estimate the economic
values of goods with the actions. The benefits of management actions in "monetary terms"
is the difference between the economic values of goods with and without the actions (see
Section 3.1.2). Monetising the costs of the actions is relatively easy; in fact, in most cases,
those costs are already in monetary terms.
Step 6: Calculate the net present value of the benefits and costs
Sixth, one should calculate the net present value (NPV) of the benefits and costs of
management actions. The benefits and costs might accrue over time. To incorporate this
time factor, an analyst assesses the NPV of a stream of net benefits {NB0, ..., NBn} that arise
over time, which is computed as
n
NPV[NB ] = NBi
n
i
i=0 (1 + r )
where r is a social discount rate and NBi is net benefits--the difference between the present
value (PV) of the gross benefits and the PV of the costs--accruing in various timings
(Tietenberg, 2003, p. 24). One can easily calculate both NPV and PV using widely-used
spreadsheet programmes. The idea of this calculation is to discount future net benefits by
interest rates so that they represent today's values.
Setting the discount rates is not an easy task; there is neither a single rate to apply nor a
consensus on how to set the rates. However, for practical purposes, Boardman et al. (2006)
recommend a discount rate of 3.5 percent for most projects whose main impacts occur
within 50 years and whose financing does not "crowd out" other investments (p. 270). U.S.
EPA (2000) suggests 2 to 3 percent for the intra-generational discounting (a relatively short
term, e.g., several decades) based on historical rates of return on relatively risk-free
investments such as government bonds, which are adjusted for taxes and inflation (p. 48);
Freeman (2003) supports this recommendation (p. 199).
Considering the rates suggested by the literature, this Guideline recommends 2 to 4 percent
as a social discount rate for the cost-benefit analyses of environmental management actions.
The Guideline also recommends conducting sensitivity analyses with respect to the discount
rate. For more information about the sensitivity analyses, see Step 7 below.
In the given example, suppose that the benefits of the management actions as well as the
costs of them accrue in various timings as described in Table 10. It is assumed that the
annual economic value of increased marine wildlife, coastal fish stock, and beach tourists
would be USD 6,450 thousand, USD 300 thousand, and USD 1,200 thousand, respectively,
following the examples discussed in this Guideline. (See Section 2.3.2.2 and Section 3.1.1
for how to estimate the increase in the economic value.) For example, the increase in the
value of wildlife value accrues from the first year soon after taking the actions, while the
value of coastal fish stock accrues from the fourth year; there is a time-lag before any effect
26
of the actions on the fish stock is seen. It is plausible to assume that the management
actions do not immediately affect "external" goods such as fish stock and beach tourism.
(For details about externalities, see Section 2.2.) The total benefit (Column 2, Table 10) is
the sum of the increased economic values, while the total cost (Column 1) is the sum of
direct costs and opportunity costs. The opportunity costs are assumed here to be the
forgone future benefits from industries that would be established if the reclaimed land area
were not reduced. Supposedly, it would take one year to establish the proposed industries;
therefore, the forgone future benefits from them would accrue from the second year and
onwards. The net benefit is the difference between the total benefit and the total cost
(Column 3 and 4).
Table 10: Benefits of management actions from a hypothetical case
Cost
Benefit
Net
benefit
Year Direct Oppor-
Total
Marine
Fish
Beach
Total
Undis-
Dis-
cost
tunity
cost
wildlife
stock tourists benefit counted counted
cost
(r=3%)
(1)
(2)
(3)
(4)
0 1,000 0
1,000 0
0
0
0
-1,000
-1,000
1
1,000
0 1,000 6,450
0
0
6,450
5,450 5,291
2 1,000
7,500
8,500
6,450
0
1,200
7,650
-850
-801
3 1,000
7,500
8,500
6,450
0
1,200
7,650
-850
-778
4 1,000
7,500
8,500
6,450
300
1,200
7,950
-550
-489
5 1,000
7,500
8,500
6,450
300
1,200
7,950
-550
-474
6 500
7,500
8,000
6,450
300
1,200
7,950
-50
-42
7 500
7,500
8,000
6,450
300
1,200
7,950
-50
-41
8 500
7,500
8,000
6,450
300
1,200
7,950
-50
-39
9 500
7,500
8,000
6,450
300
1,200
7,950
-50
-38
10 500
7,500
8,000
6,450
300
1,200
7,950
-50
-37
Total 8,500
67,500
76,000
64,500
2,100
10,800
77,400
1,400
1,552
Unit: USD thousand
The total net benefits are different depending on whether they are discounted or not. In this
example, both the net benefits (discounted and undiscounted) are positive. Discounted with
the 3-percent interest rate, the NPV is approximately USD 1.6 million. In other words, the
management actions are preferable according to the decision criteria discussed in Section
3.1.4.
Step 7: Conduct sensitivity analyses
Seventh, one should conduct sensitivity analyses not only to incorporate uncertainties but
also to check the robustness of analytical results. There might be uncertainties about the
impacts--benefits and costs--of management actions, that were predicted in Step 4, or
about the discount rates used in Step 6. To incorporate the uncertainty with respect to the
discount rates, an analyst should recalculate net benefits, using different rates. If net
benefits still remain positive (or negative), one can be confident about supporting (or
opposing) the proposed management actions.
For example, consider using different discount rates that are either slightly higher or lower
than the original 3-percent discount rate. Table 11 shows estimated discounted net benefits
or NPVs in the example with the following three different rates used: 1, 3, and 5 percent. In
this example, the signs of net benefits for all three discount rates are positive. That is, an
27
analyst can conclude with confidence that the proposed management actions make sense
economically.
Table 11: Sensitivity-analysis results: Net present value of management actions from
a hypothetical case
Net present value
Year
r = 1%
r = 3%
r = 5%
0 -1,000
-1,000
-1,000
1 5,396
5,291
5,190
2 -833
-801
-771
3 -825
-778
-734
4 -529
-489
-452
5 -523
-474
-431
6 -47
-42
-37
7 -47
-41
-36
8 -46
-39
-34
9 -46
-38
-32
10 -45
-37
-31
Total
1,455
1,552
1,632
Unit: USD thousand
Step 8: Make recommendations
Lastly, one should prepare recommendations based on the results of cost-benefit analyses.
Following the decision criteria discussed in Section 3.1.4, an analyst should recommend that
decision-makers adopt management actions with a positive NPV (or with the largest NPV),
or dismiss the actions with a negative NPV (or with small NPVs). Explaining the
methodology and data processing used in the analyses, the analyst should also present (as
displayed in Tables 10 and 11) the flow of benefits and costs in addition to a summation of
values (i.e., NPV) (U.S. EPA, 2000, p. 48). That would provide decision-makers with an
opportunity to examine the validity and reliability of an estimated NPV(s).
28
4 Summary
By analysing the benefits and costs of an action or a group of alternative actions, economic
analyses help in deciding whether to implement a specific management action or deciding
what management actions should be implemented in order to address environmental
problems. As a result, environmental decisions or management actions will be efficient.
To measure the economic value of environmental goods, which is defined as the sum of
consumer surplus and producer surplus, various techniques are available, including
empirical technique, TCM, and CVM. The selection of the techniques depends on the
characteristics of goods to be evaluated. If concerned goods are traded in the market, their
market prices and trading volumes are useful to estimate the value of the goods. The
empirical technique takes this approach. If the goods are not traded in the market, however,
either the market information of relevant goods or the information of consumer preference
surveyed for the goods would be applicable to estimate their value. A typical example of the
former approach is the TCM; meanwhile, that of the latter is the CVM. The TCM uses the
information on how much people spend to consume environmental goods as a proxy
variable for their economic value. The CVM uses survey results on individuals' willingness
to pay for the goods in order to calculate their value.
Cost-benefit analyses examine the economy or efficiency of a management action(s). The
analyses compare the net benefits--the difference between gross benefits and costs--of
management actions under two scenarios: with or without the actions. By definition, the
gross benefits of the actions are the prevented future loss measured in economic value. The
costs of the actions are the costs incurred to implement proposed actions. Given analytical
results, it is logical to accept the proposed actions if the net benefits are positive, or to
decline the actions if the net benefits are negative.
The procedure of a cost-benefit analysis consists of eight steps, including calculating the
NPV of the gross benefits and costs as well as conducting sensitivity analyses. To calculate
the NPV, it is recommended to use 2 to 4 percent as a social discount rate. It is also
recommended to conduct sensitivity analyses with respect to the discount rate in order not
only to incorporate uncertainties but also to check the robustness of analytical results. If the
net benefits still remain positive (or negative), as a result of the sensitivity analyses, one can
be confident about his/her conclusion to support (or oppose) the proposed management
actions.
29
References
Abt Associates Inc. (Ed.). (2005). Benefits transfer and valuation databases: Are we
heading in the right direction? Washington, DC: U.S. EPA.
Boardman, A. E., Greenberg, D. H., Vining, A. R., & Weimer, D. L. (2006). Cost-benefit
analysis: Concepts and practice. (3rd ed.). Upper Saddle River, New Jersey: Pearson
Prentice Hall.
Freeman, A. M. III. (2003). The measurement of environmental and resource values:
Theory and methods. (2nd ed.). Washington, DC: Resources for the Future.
Grigalunas, T. A., Opaluch, J. J., Diamantides, J., & Brown, G. M. (1995). Environmental
economics for integrated coastal area management: Valuation methods and policy
instruments. Nairobi, Kenya: UNEP.
Gujarati, D. N. (1995). Basic econometrics. (3rd ed.). New York: McGraw-Hill.
Kim, D. (2003). A bioeconomic analysis of the management policies for the United States
Gulf of Mexico red grouper fishery. Ocean and Polar Research, 25 (4), 483-491.
Lipton, D. W., Wellman, K., Sheifer, I. C., & Weiher, R. F. (1995). Economic valuation of
natural resources: A handbook for coastal resource policymakers. Washington, DC:
U. S. Department of Commerce & National Oceanic and Atmospheric Administration.
Loomis, J. B. (1988). Contingent valuation using dichotomous choice models. Journal of
Leisure Research, 20 (1), 46-56.
Markandya, A., Harou, P., Bellu, L. G., & Cistulli, V. (2002). Environmental economics for
sustainable growth: A handbook for practitioners. Cheltenham, UK: Edward Elgar.
Markandya, A., Perelet, R., Mason, P., & Taylor, T. (2001). Dictionary of environmental
economics. Sterling, VA: Earthscan.
Mitchell, R. C., & Carson, R. T. (1989). Using surveys to value public good: The contingent
valuation method. Washington, D. C.: Resources for the Future.
Pagiola, S., Ritter, K., & Bishop, J. (2004). How much is an ecosystem worth?: Assessing
the economic value of conservation. Washington, DC: World Bank.
Pindyck, R. S., & Rubinfeld, D. L. (1995). Microeconomics. (3rd ed.). Englewood Cliffs,
New Jersey: Prentice Hall.
Taromaru, H. (2005). Categorical data analysis for socio-human sciences. Kyoto, Japan:
Nakanishiya.
Tietenberg, T. (2003). Environmental and natural resource economics. (6th ed.). Boston,
Massachusetts: Pearson Education.
U.S. Environmental Protection Agency. (2000). Guidelines for preparing economic
analyses. Washington, DC: U.S. Government.
Yellow Sea Large Marine Ecosystem Project. (2000). YSLME Preliminary Transboundary
Diagnostic Analysis. UNDP/GEF.
Yellow Sea Large Marine Ecosystem Project. (2005a). Reducing environmental stress in
the YSLME: Report of the first meeting of the regional working group for the pollution
component. UNDP/GEF.
Yellow Sea Large Marine Ecosystem Project. (2005b). Reducing environmental stress in
the YSLME: Report of PSC and technical meetings. UNDP/GEF.
Additional reference (website)
Ecosystem Valuation. http://www.ecosystemvaluation.org/.
30
Appendixes
Appendix 1: Consumer surplus estimated by market demand curve for the site
The consumer surplus for the recreational site in the example can be calculated by the
following two steps (Boardman et al., 2006, pp. 358-359):
(1) Derive the market demand curve for the site, using the representative individual's
demand curve; and
(2) Calculate the area under the estimated demand curve and above the admission fee.
First, according to the individual's demand curve estimated in Section 2.3.2.1, the number of
visits per person per year would decrease by 0.2 as V = 19 0.2TC (the inverse of TC = 95
5V). That is, a ten dollar increase in travel cost per visit (TC), for example, would reduce
the number of visits (V) by 2. Since the current admission fee is USD 10, if the fee were
USD 20, the number of visits from Zone A would be 13 (15 2 = 13). Similarly, the number
of visits from Zone B, Zone C, and Zone D would be 11, 4, and 1, respectively. The number
of visits from Zone E (1 2 = -1) would be considered as 0 because negative numbers of
visits are not possible. With the number of visitors by zone considered, the market demand
for the site with the 20 dollar admission fee would be 330,000 visits per year (13 x 10,000 +
11 x 10,000 + 4 x 20,000 + 1 x 10,000 + 0 x 10,000 = 330,000). Table 12 shows the market
demand for the site with various admission fees. In this example, it is assumed that the fee
be raised by 10 dollars. Figure 9 shows the estimated market demand curve based on the
information of the fees and the visits described in Table 12.
Note that there would be no demand for the site (i.e., there is no visit) if the admission fee
were USD 85. According to the individual's demand curve, the number of visits would be 0
when the travel cost per visit were USD 95. (See Figure 4.) Recall that the travel cost per
visit includes the admission fee which is currently USD 10. That is, the representative
individual would be willing to pay USD 85 at a maximum for the site (95 10 = 85).
31
Note: * Thousand per zone
To
E
D
C
B
A
Z
one Popu
t
a
l
Admission
10,00
10,00
20,00
10,00
10,00
latio
fee
0
0
0
0
0
n
perso
Per
Annu
13
15
1
3
6
n
al visits
10
zone
Per
440
120
130
150
10
30
*
perso
Annu
Per
Table 12: Market demand for a h
11
13
0 0
1
4
n
al visits
20
zone
330
110 9
130
Per
10
80 2
perso
Annu
Per
11
0 0
0 0
n
al visits
30
zone
240
110
Per
40 0
90 7
perso
Annu
Per
32
0 0
0 0
9 90
n
al visits
40
zone
y
160
Per
pothetical recreational site
70
0
perso
Annu
Per
0 0
0 0
0
5
7 70
n
al
50
visits
zone
120
Per
50 3
0 0
perso
Annu
Per
0 0
0 0
5 50
n
al visits
60
zone
Per
80
30 1
0 0
perso
Annu
Per
0 0
0 0
3 30
n
al visits
70
zone
Per
40
10 0
0 0
perso
Annu
Per
0 0
0 0
1 10
n
al visits
80
zone
Per
10
0 0
0 0
perso
Annu
Per
0 0
0 0
0
n
al visits
85
zone
Per
0
0
0
0
90
a 0 ,
( 85
)
80
70
) 60
SD
U 50
e
e
(
o
n
f
d
b
40
160 ,
( 4
0
)
i
ssi
d
m 30
A
20
e
f
c
10
440
( ,
10
)
0
0
50
100
150
200
250
300
350
400
450
500
Number of visitors per year (thousands)
Source: Boardman et al., 2006, p. 359
Figure 9: Estimated market demand curve for a hypothetical recreational site
Second, the market consumer surplus can be measured by calculating the area under the
estimated market demand curve and above the original admission fee (i.e., Area abce). For
simplicity, suppose that the demand curve is linear between points a and b and between
points b and c; then, the market consumer surplus for the site is estimated at USD 12.6
million per year as follows.
Consumer surplus for a hypothetical recreational site
= Area abce
= Area abd + Area dbfe + Area bcf
= (85 - 40)×160×1 2 + (40 -10)×160 + (40 -10)×(440 -160)×1 2
= USD 12,600 thousand per year
In theory, the market consumer surplus estimated above should be the same as that
estimated in Section 2.3.2.1. However, in this Guideline, the estimation in this section (USD
12.6 million) is more than that in Section 2.3.2.1 (USD 11.9 million); the former
overestimates the market consumer surplus. The reason for the overestimation is that it is
assumed in the above calculation in this section that the market demand curve consists of
simply two straight line segments (i.e., line ab and line bc). If the area were summed with
more segments assumed, the sum would be USD 11.9 million (D. L. Weimer, personal
communication, March 30, 2008).
33
Appendix 2: Logit model
The logit model is defined as:
P
i
L = ln
= + X
i
(1- P )
1
2
i
i
where Pi / (1 Pi) is the ratio of the probability that an event occurs (e.g., respondents would
be willing to pay or reply "yes" in the example in Section 2.3.2.2) to the probability that it
does not occur; this ratio is called the "odds ratio." L, called the logit, is the log of the odds
ratio (Gujarati, 1995, p. 555). X, an explanatory variable, represents bid prices in the
example, while 1 and 2 are coefficients. Taking the exponential of L gives:
exp(L)
Pi
= expln
(
= exp + X
1- P
1
2
i )
(
i )
Pi
(
= exp + X
1 - P
1
2
i )
(
i )
exp( + X
1
2
i )
P =
i
1+ exp( + X
1
2
i )
where Pi is, as defined above, the probability that respondents would reply "yes" to given bid
prices, X , in the example (Taromaru, 2005, p. 176).
34
Appendix 3: Individuals' consumer surplus estimated by CVM
The individuals' consumer surplus (i.e., the area under the estimated logistic regression
function) in the example can be calculated by:
(1) Dividing the range of bid prices into equal segments of width n (e.g., n = 1) (Column
1 in Table 13);
(2) Calculating the probability of acceptance at each of these points, using the estimated
logistic regression function (Column 6);
(3) Finding the average acceptance value for adjacent pairs of points (Column 7);
(4) Multiplying each of these averages by n (i.e., n = 1) (Column 7); and
(5) Summing all these products to get the estimate of the area (See the last row in
Column 7).
According to the calculations, the area or the estimated individuals' consumer surplus for the
site is approximately USD 21.5.
Table 13: Estimated individuals' consumer surplus for coastal site rehabilitation
Bid price
1+exp
p^=exp(*)/
Response logit(p) exp(logit(p))
Area
(USD)
(logit(p))
(1+exp[*])
(1) (2) (3) (4) (5) (6) (7)
0 3.321
27.688
28.688
0.965
0.962
1 3.165
23.689
24.689
0.959
0.956
2 3.009
20.267
21.267
0.953
0.949
3 2.853
17.340
18.340
0.945
0.941
4 2.697
14.835
15.835
0.937
0.932
5 1 2.541
12.692
13.692
0.927
0.921
6 1 2.385
10.859
11.859
0.916
0.909
7 1 2.229
9.291
10.291
0.903
0.896
8 2.073
7.949
8.949
0.888
0.880
9 1 1.917
6.801
7.801
0.872
0.863
10 1 1.761
5.818
6.818
0.853
0.843
...
...
...
...
...
...
...
55 0 -5.259
0.005
1.005
0.005
0.005
56 -5.415
0.004
1.004
0.004
0.004
57 -5.571
0.004
1.004
0.004
0.004
58 -5.727
0.003
1.003
0.003
0.003
59 -5.883
0.003
1.003
0.003
0.003
60 0 -6.039
0.002
1.002
0.002
Total area
21.501
Note: The log of the odds ratio (Column 3) is calculated, using the estimated individuals'
WTP function (i.e., RY = 3.321 0.156BP; see Section 2.3.2.2 in this Guideline).
35