Lecture 7 - Valuing Environmnetal Benefits - State Preference Method

THAI NGUYEN UNIVERSITY OF AGRICULTURE AND FORESTRY
ADVANCED EDUCATION PROGRAM (AEP)
Natural resource economics
Lecture 7: Valuing environmental Benefits:

Stated Preference Method
Contingent valuation method (CVM) & Choice experiment
Do Xuan Luan, PhD

Faculty of economics and rural development, TUAF
17/09/2022 Do Xuan Luan, PhD, Ferd//TUAF 1

List of all lectures
1. Introduction
2. Externalities
3. Market failure, its causes and suggestions to correct
4. Economics of Pollution
5. Environmental costs
6. Valuing environmental Benefits- Revealed Preference Method
7. Valuing environmental Benefits- State Preference Method
8. Cost-Benefit Analysis (CBA)
9. Interventions of Government
10. Review & discussions
17/09/2022 Do Xuan Luan, PhD, Ferd/TUAF 2

Learning outcome
After this lecture, students should be able to:
1. Explain the main idea of using the Contingent Valuation Method

(CVM) to value non-marketed resources, such as recreation,
wildlife, and environmental quality.
2. Explain the main idea of using the Choice experiment to value

non-marketed resources.
3. Understand the Pros and Cons of each approach
4. Be able to measure benefits through a simple example.

Stated Preference Method is applied to measure non-marketed resources, such as
recreation, wildlife, and environmental quality.
Recreation
Wildlife Environment

Review: Two approaches to measure the benefits of environmental
quality
Aversion costs
Revealed Travel cost

preference method
approach
Hedonic pricing
techniques
Measure
environmental
benefits
Contingent
valuation
Stated method (CVM)
preference
techniques
Choice
experiment

Review: Revealed preference approach vs. Stated
preference approach
• Revealed preference approach – infer the value of environmental
goods from other market transactions
• Note that revealed preference approaches get at use values, but not non-
use values.
• Stated preference approach which we are going to discuss in this
lecture – ask individuals hypothetical questions about their
willingness to pay or willingness to accept.

STATED PREFERENCE METHOD
• Stated preference method can be applied to any situation to determine

willingness to pay (WTP) or their willingness to accept (WTA) for a
hypothetical scenario.
• Allows collection of information about respondent preferences for the

environmental amenities of interest by observing choices in
hypothetical situations presented in a survey.
• Ask individuals hypothetical questions about their WTP/WTA;
• Use survey techniques to elicit WTP/WTA for a marginal improvement or

for avoiding a marginal loss;
• WTP/WTA reflect the total economic value that respondents place on a

resource, including use and nonuse benefits.
Stated Preference Approaches
Stated
preference
approach
1. Contingent 2. Choice
valuation method experiment
Respondent asked to
(CVM) consider an
Use surveys to assess environmental
WTP/WTA commodity defined by
to bring about a several attributes.
specific Choose among three
environmental or more options
during each task
improvement.

Contingent valuation method (CVM): How does it work?
CVM values the environmental quality and directly ask questions about
willingness to pay (WTP) to have an improvement in environmental
quality, or willingness to accept (WTA) to abstain from having an
improvement in environmental quality.
• Maximum willingness to pay (WTP) for an improvement in

environmental quality. E.g: WTP to avoid a worsening in environmental
quality.
• Minimum willingness to accept (WTA) to abstain from an improvement

in environmental quality. E.g: WTA to accept a worsening in
environmental quality.

Contingent Valuation Method (CVM)
• An economic tool that uses surveys to question people regarding their

willingness to pay for a good, such as the preservation of hiking
opportunities or air quality.
• CVM questions can be phrased in terms of either WTA, for a scenario that
decreases utility, or WTP, for a scenario that increases utility.
• Observed choices are contingent on scenarios posed in the survey and the
environmental economics literature commonly uses the term contingent
valuation (CV) to describe the process of utilizing stated preference data for
valuation.
• CVM is an inherently more flexible tool than revealed preference (RP);
• CVM is also needed for cost‐benefit analysis;

Six main component of a successful CVM study
• Contingent Valuation- Uses survey to access individual’s reaction on
changes in environmental quality.
• It analyze how much are they willing to pay in a scenario.
• As Carson (1991) noted, there are six main component of a successful
CV study. These are
1. Define Market Scenario
2. Choosing Elicitation Method
3. Design market administration
4. Sample Design
5. Experimental Design
6. Estimation of Willingness to Pay Function

CONTINGENT VALUATION METHOD: Define Market Scenario
• Is the information to be conveyed to a respondent? (i.e. one who

will be asked about willingness to pay);
• To place the respondent in the right time frame of CONTINGENT

VALUATION METHOD mind to give meaning response to
questions;
• Description of the market should be realistic to the respondent and

Defining appropriate payment mechanism.

CONTINGENT VALUATION METHOD: Sample, Experiment
design and estimation of WTP
• Sample Design: It involves two steps: First, select the group

(relevant for the study) from which to draw the sample. Second,
draw the random sample;
• Experimental Design: Experimental design requires careful design

of survey instrument, its administration and its ultimate statistical
analysis.
• Estimation of the function of Willingness to Pay:

The last step of the Contingent Valuation Method is to take the
survey results and correctly estimate the WTP functions.

CONTINGENT VALUATION METHOD
1. Define Market Scenario

To place the respondent in the right time frame to give meaning response to
questions.
2. Choosing Elicitation Method

To decide how best to obtain the valuation process. There are four ways of eliciting
value: direct question, bidding game, payment card and referendum choice/or
discrete choice.
3. Design Market Administration

Three approaches to survey administration: mail, telephone and in-person.
4. Sample Design
It involves two steps: First, select the group (relevant for the study) from
which to draw the sample. Second, draw the random sample
5. Experimental Design
Experimental design requires careful design of survey instrument, its
administration and its ultimate statistical analysis.
The last step of the Contingent Valuation Method is to take the survey
6. Estimation of Willingness to Pay Function:
results and correctly estimate the WTP functions.
15
17/09/2022 Do Xuan Luan, PhD, Ferd//TUAF
CONTINGENT VALUATION METHOD:
4 ways of eliciting value
• The next step is to decide how best to obtain the valuation process.
• There are four ways of eliciting value:
1) direct question,
2) bidding game,
3) payment card and
4) Referendum or discrete choice.

Direct question
Ask people their willingness to pay (WTP) to bring about a specific
environmental improvement.
 The problem must be described carefully.
 It must be specific
 Types of questions:
• Open-ended: ask respondent for maximum WTP. How much are you
willing to pay for…………..?
• Close-ended: ask respondents whether they are WTP a certain
amount. This amount is varied across respondents. Are you willing
to pay…..$ for…………?
• However, this suffers from a great demerit in the sense that there are few real
markets in which we ask the respondent to generate data;
• And in most occasion people may not spend much effort in determining their
willingness to pay which may result in extreme response (either zeroes and
very large numbers)
Example of Close-ended question:
Single Bounded
One question;
• Are you willing to pay….$B for non-market good i? Yes, No.
How to estimate B in single-bounded case?
(1) We run the probit model for wtp1 as a function of respondents’ characteristics
such as income, age, education….
(2) Calculate mean and median for WTP.
(see Krinsky and Robb approach)
. wtpcikr bid1 income age marriage edu children, reps(5000)
Krinsky and Robb (95 %) Confidence Interval for WTP measures (Nb of reps:
5000)
MEASURE WTP LB UB ASL* CI/MEAN
MEAN/MEDIAN 9.28 -70.51 61.00 0.1894 14.18
*: Achieved Significance Level for testing H0: WTP<=0 vs. H1: WTP>0
LB: Lower bound; UB: Upper bound

Evaluating contingent valuation: Open-ended/Close-ended
• Content validity: do surveys lead to true values? Is best practice being

followed?
• True zero
1. Can not afford (true zero)
2. Government business (protest zero)
3. Don care about environment (not important): marginal utility is zero (true
zero)
4. Do not know what is the meaning of the question (protest zero also)
5. Other reasons (i pay tax, the pollutants should pay; Gov not transparent, no
trust Government, money is not used for environment…)- protest zero also
• Protest zeros (usually thrown out by researchers)
• Construct validity: are stated preference estimates consistent with theoretical

predictions?
• Criterion validity: do stated preference estimates match real payment?
• Convergent validity: are stated and revealed preference estimates the same?

While early studies used a range of method, standard practice now
is to use close-ended responses.
An example: use a contingent valuation survey of improved water

quanlity.
Please answer the following questions honestly.
1. Are you willing to pay $ _ per year to improve the water quality of
Hoan Kiem lake? YES NO
2. What is the maximum your household would be willing to pay per

year to preserve our lake? $________
If your answer to #2 was zero, please respond to question #3; otherwise

answer #4.

An example: use a contingent valuation survey of
improved water quanlity
3. Why would you not contribute to this program? (Please check only
one.)
□ This program is not important to me.
□ I am on a limited budget.
□ I do not think that this program would be effective.
□ I do not think that I should be responsible for this program.
□ I think the current system of water management is adequate.
□ I do not support any new government programs.
□ Other: ________________________________________________
Please skip question #4.

An example: use a contingent valuation survey of improved
water quanlity
4. Why would you pay the amount in #2?

□ This program is important to me.
□ I think it is our responsibility to protect our water.
□ I want to contribute to a good cause.
□ Other: _________________________________________________

quanlity
PERSONAL INFORMATION
This information is for statistical purposes only and is completely anonymous.
1. □ MALE □ FEMALE
2. AGE: _______ YEARS
3. HOW LONG HAVE YOU LIVED IN VIETNAM? ____ YEARS
4. DO YOU:
□ OWN YOUR HOME
or □ RENT
5. IN THE LAST YEAR HOW OFTEN DID YOU VISIT THE LAKE?
□ 0 visits
□ 1 – 10 visits
□ over 10 visits

quanlity
PERSONAL INFORMATION
6. PLEASE CHECK THE HIGHEST GRADE LEVEL YOU HAVE SUCCESSFULLY
COMPLETED.
□ ELEMENTARY
□ HIGH SCHOOL
□ SOME COLLEGE
□ COLLEGE DEGREE
7. ABOUT HOW MUCH WAS YOUR HOUSEHOLD INCOME?
□ Less than $17,500
□ $17,500 to $30,000
□ $30,000 to $42,500
□ $42,500 to $60,000
□ Over $60,000
Thank you for your cooperation. Please return the survey in the enclosed envelope
at your earliest convenience.
• 17/09/2022 Do Xuan Luan, PhD, Ferd//TUAF 24
Choosing Elicitation Method: Bidding Game
• Bidding game approach was first used by Randal et.al (1974). This
approach involves a WTP number and seeks a yes-no response.
• If the respondent replies yes, the amount is gradually increased until a

no response is received.
• Similarly, if the respondent replies no, the amount is gradually

decreased until a yes is received.
• The main problem with this approach is the starting-point bias.

Bidding Games
• Biding games: ask respondents whether they are WTP a certain

amount. If they say yes, ask them about a higher amount, until you
find the highest amount they are WTP.
• Initially, a set of bidding format was used to generate the amount of

Willingness To Pay (WTP) for conserving a recreational place:
• 1st bid = US$ 0.50
• 2nd bid = US$ 1
• 3rd bid = US$ 8

Payment Card
• The payment card is a willingness to pay (WTP) elicitation approach,

initially developed by Mitchell and Carson (1989) to address survey
bias in evaluating WTP toward public environmental and resource
projects.
• The payment card approach, unlike the referendum type of

questions, allows respondents to choose the amount that best reflects
the maximum price they would like to pay for a product.
• A card with a number of figures, spanning the range of responses

that might be expected.
• Each card has payment amounts along with several reference

expenditure amount.

Payment Card
Important notes:
• A payment mechanism must be specified and believable.
• e.g. reduce pollution from steel production to improve sea quality.

• Literature review and trial surveys should be used to design the
intervals of choices
• Information about the respondents (age, income, education…) should

be gathered to allow the verification of results, estimation of income
elasticities…

Payment Card:
Which of the amounts below most closely indicates your maximum
willingness to pay annually, for the conservation of Phong Nha-Kẻ Bàng
National Park?
____$10,000 ____$40,000 ____$70,000

____$20,000 ____$50,000 ____$80,000
____$30,000 ____$60,000 ____$90,000

$0
Payment card: Example 2 for food policy
How much are you willing to pay for one liter of organic milk higher than
one liter of conventional milk?
a) Less than 0.2$
b) 0.2-0.5$
c) 0.5-0.7$
d) More than 0.7$
How to design those intervals? Otherwise (strategic bias).

Two common methods:
(1) Literature review
(2) trial surveys (30-50 samples), look at frequency distribution, mean &
median. The idea is to design intervals such that the frequency
distribution is normal distribution.

Payment card: Estimation techniques
How to estimate WTP? Econometrics models
(1) OLS: take the medium level of each interval. . However, this
method is not widely used due to many errors or bias,.
(2) Interval Regression: widely used.
(3) Bivariate probit model

Dichotomous choice
• Dichotomous choice is presented as a branching question between
two things that are opposite or entirely different.
• Respondent is asked a series of dichotomous choice questions until
some point estimate of willingness to pay (WTP) is reached;
• If the individual responds "yes" to the first bid, the second bid is some
amount higher than the first bid; if the individual responds "no" to the
first bid, the second bid is some amount lower than the first bid

Example of Discrete Choice for CVM
Idea:
Respondents will be asked:
1. Are you willing to pay B$ for non-market good i (e.g: organic vegetable, blue
sky,….)? Yes, No?
2. If Yes, are you willing to pay B$ + k for non-market good i?
3. If No, are you willing to pay B$-k’ for non-market good i?
This figure shows that this method is quite close to Payment Card method, using Interval
regression. How to estimate?
(1)bivariate probit model
(2) Bivariate probit withImpose constraints (Impose behavior constraints)

Discrete Choice Format for CVM: Estimation in stata Software
biprobit (wtp1 bid1 income age gender marriage edu children) (wtp2 bid2 income age
gender marriage edu children)

Discrete Choice for CVM: Estimation in Stata statistical Software
After anlyzing determinants of WTP1 và WTP2, we estimate WTP1 và WTP2.
(And calculate the WTP1 and WTP2)

CVM- A summary
Take note that, there are 4 ways for bidding mechanisms, which are:
1. open-ended,
2. payment card,
3. bidding games, (bidding game works by asking a sequence of questions
until maximum is found)
4. and lastly, the closed-ended or the discrete choice. Single and double –
bound CVM are closed-ended methods , which use straightforward
questions for bidding mechanisms.
1 2 4
Open-ended or Payment card Close-ended or
Direct question Discrete choice

Contingent Valuation Method (CVM)

Discussion: Single – bound vs. Double – bound
• Single-bound just requires less information, so it is easier to
implement especially in data collection and estimation stages
• On the other hand, it is well known that the double bound is more
efficient and less biased estimates than the single bound estimator.

Summary: choice of methods
Which one: open-ended, Payment Card and Discrete Choice, which is better?
1. open-ended: no bias (because we have no zero observations, no hint).
2. Discrete Choice: initial bidding value is very important. Changes in the

initial bidding value might change behavior.
3. Payment Card: Design of question is very important. To find the value of

bidding:
(1) literature review: mean and median of bidding (B)
(2) trial survey to find out the distribution choices:
So, in term of bias from design, open-ended method might be better.
In term of flexibility, discrete choice format might be better (not many zero
observations, few problems of econometrics).
Summary: choice of methods
Are the estimated WTP values different?
WTP obtained from three methods (open-ended, payment card, discrete

choice) different?
• WTP_open.ended ≤ WTP_discrete choice method (because of zero

observations).
• In general, in practice, we often combine two methods together.
• Discrete choice (Single bounded, double bounded);
• Different methods should get close results, otherwise we have

problems of design problems.
• Combination of methods to make results more robust.

Choice experiment
• Example: meauring WTP for organic beef. Each attribute consist of some
characteristics.
• Assume that we design cards with 4 attributes (lean, red, organic, price) and
offer 3 options for customers: A, B and C (not buy). Each option is one type of
organic beef.
• For instance, the attribute “lean” include characteristics “large”, “medium”

and “small”. Attribute red has characteristics such as red or dark…..
• Assume that the consumer select B. He/she prefer beef with lean (medium),
red, conventional, price 9euro/kg.
• We need to translate data from cards distributed to each respondent into excel.
• How to design cards and how to include data into file excel?
Choice experiment
We incorporate this card into excel file as follows:
Choice Lean red organic price

0 (not select A) 2 0 1 5
1 (select B) 1 1 0 9
0 (not select C). opt-out 0 0 0 0

- do not want to buy

Choice experiment
Assume that the consumer select A. He/she prefer beef with lean
(large), dark, organic, and price 5euro/kg. We will incorporate the
option from this customer into excel as follows:

1 (select A) 2 0 1 5
0(not select B) 1 1 0 0
0 (not select C) 0 0 0 0

Choice experiment
We incorporate this card into excel file as follows:

0 (not select A) 2 0 1 5
1 (select B) 1 1 0 9
0 (not select C). 0 0 0 0

opt-out - do not
want to buy

Choice experiment
• Option C (no purchase), we can futher ask which type of product that
the consumer is currently use and then we code and import in excel
like cases of option A and B. For example, the consumer currently uses
beef with characteristics (lean-medium, colour-red, conventional and
price of 8).
• So, for each customer interviewed, we need to distribute him/her 12

cards (C48, ???) . Each card, the consumer has three options A, B and C.
• If we make interviews with 100 customers, number of cards should be:

100*12=1200. In other words, we will have 1200 observations in this
case. Each card represents for a type of beef with typical characteristics.

How to design cards and how to include data into file excel?

Choice experiment: Estimation in Stata
• Analye determinants of choice and WTP for each attribute of beef
product

Choice experiment: estimation in Stata
Interpretation:
• Calculate WTP for each attribute of beef products (actually, we estimate
WTP after chocie estimation)
• Results show that leanh, leanm, red and organic have an impact on WTP
• Results tell us the willingess to pay for each attribute. For example, wtp
of leanh=2.35 meaning that if a type of beef with “lean”, will increase
WTP by 2.35 euros.

Choice experiment: estimation in stata
• Because respondents are different, groups are different. We need to use
fixed effect to control. Therefore, we use clogit

Estimate WTP
Compare WTO for attributes of this estimation with previous results from normal logit
estimation. By 1000 replications

Choice experiment: Estimation in Stata
• Increase price decreases utility

• To control for differences between respondents and groups, we use mixlogit
(control for group differences and respondent differences and all coefficients are
random)

• Calculate WTP

Choice experiment
• If we consider all cards with no purchase options, the procedures
of analysing determiants of choice and estimating WTP will be as
follow:

Choice experiment: 2nd example:A Survey on Maine Residents
About their Preferences for Alternative Forestharvesting Practices
The state of Maine was considering purchasing a 23,000 acre tract of forest
land to manage.
Attributes used in the survey included the number of live trees,
management practice for dead trees, percent of land set aside, and a tax
payment. Three levels of each management attribute and 13 different tax
prices were considered. Table 4.2 reproduces the attributes and levels.

Choice experiment: 2nd example:A Survey on Maine Residents About their
Preferences for Alternative Forestharvesting Practices
Residents were given a choice set of four different alternative management plans and the
status quo (no purchase). Table 4.3 demonstrates a sample survey question. This type of
survey has evolved from both contingent valuation and marketing studies.
This approach allows the respondent to make familiar choice (choose a bundle) and allows
the researcher to derive marginal willingness to pay for an attribute from that choice.

Some conclusions
1. Stated preference approach– ask individuals hypothetical questions about their
willingness to pay or willingness to accept.
2. Contingent valuation method (CVM) and Choice experiments are two

techniques of stated preference approach.
3. CVM includes techniques such as open-ended/close-ended, Payment Card

and Discrete Choice;
4. Choice experiment approach is used to meaure WTP based on attributes

(consist of some characteristics) which contribute to the value of the
product/service.
5. Each approach has pros and cons. The choice of each approach depends on our
arguments.

Potential studies
1. Willingness to pay for safe (or organic) vegetables in Vietnam: An
empircal analysis in Thai Nguyen city.
2. Willingness to pay for free-additive mooncake (Trung cake, square
cake…) in Vietnam;
3. Using local knowledge to model asymmetric preference formation in
willingness to pay for environmental services: A case study of Ba Be
National Lake, Bac Kan province, Vietnam.
4. Tourists’ willingness to pay for biodiversity conservation and
environment protection, Ba Be Lake protected area: Implications for
entrance fee and sustainable management;
5. Household’s willingness to pay for arsenic safe drinking water in
Vietnam;
6. Willingness to pay for preferential loans in Vietnam: An case study of
preferential credit of Vietnam Bank for Social Policy (Attribute of credits,
choice experiment design,….).
www.themegallery.com

Lecture 7 - Valuing Environmnetal Benefits - State Preference Method

Uploaded by

Copyright:

Available Formats

Lecture 7 - Valuing Environmnetal Benefits - State Preference Method

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Lecture 7 - Valuing Environmnetal Benefits - State Preference Method

Uploaded by

Copyright:

Available Formats

THAI NGUYEN UNIVERSITY OF AGRICULTURE AND FORESTRY

ADVANCED EDUCATION PROGRAM (AEP)

Natural resource economics

Lecture 7: Valuing environmental Benefits:

Do Xuan Luan, PhD

17/09/2022 Do Xuan Luan, PhD, Ferd//TUAF 1

3. Market failure, its causes and suggestions to correct

6. Valuing environmental Benefits- Revealed Preference Method

7. Valuing environmental Benefits- State Preference Method

8. Cost-Benefit Analysis (CBA)

10. Review & discussions

17/09/2022 Do Xuan Luan, PhD, Ferd/TUAF 2

1. Explain the main idea of using the Contingent Valuation Method

2. Explain the main idea of using the Choice experiment to value

3. Understand the Pros and Cons of each approach

4. Be able to measure benefits through a simple example.

17/09/2022 Do Xuan Luan, PhD, Ferd//TUAF 3

17/09/2022 Do Xuan Luan, PhD, Ferd//TUAF 4

Revealed Travel cost

17/09/2022 Do Xuan Luan, PhD, Ferd//TUAF 5

• Revealed preference approach – infer the value of environmental

goods from other market transactions

• Stated preference approach which we are going to discuss in this

lecture – ask individuals hypothetical questions about their

willingness to pay or willingness to accept.

17/09/2022 Do Xuan Luan, PhD, Ferd//TUAF 6

• Stated preference method can be applied to any situation to determine

• Allows collection of information about respondent preferences for the

• Ask individuals hypothetical questions about their WTP/WTA;

• Use survey techniques to elicit WTP/WTA for a marginal improvement or

• WTP/WTA reflect the total economic value that respondents place on a

17/09/2022 Do Xuan Luan, PhD, Ferd//TUAF 8

• Maximum willingness to pay (WTP) for an improvement in

• Minimum willingness to accept (WTA) to abstain from an improvement

17/09/2022 Do Xuan Luan, PhD, Ferd//TUAF 9

• An economic tool that uses surveys to question people regarding their

• CVM is an inherently more flexible tool than revealed preference (RP);

• CVM is also needed for cost‐benefit analysis;

17/09/2022 Do Xuan Luan, PhD, Ferd//TUAF 10

1. Define Market Scenario

2. Choosing Elicitation Method

3. Design market administration

6. Estimation of Willingness to Pay Function

17/09/2022 Do Xuan Luan, PhD, Ferd//TUAF 11

• Is the information to be conveyed to a respondent? (i.e. one who

• To place the respondent in the right time frame of CONTINGENT

• Description of the market should be realistic to the respondent and

17/09/2022 Do Xuan Luan, PhD, Ferd//TUAF 12

• Sample Design: It involves two steps: First, select the group

• Experimental Design: Experimental design requires careful design

• Estimation of the function of Willingness to Pay:

17/09/2022 Do Xuan Luan, PhD, Ferd//TUAF 13

1. Define Market Scenario

2. Choosing Elicitation Method

3. Design Market Administration

• There are four ways of eliciting value:

3) payment card and

4) Referendum or discrete choice.

$10,000 $40,000 $70,000