Download presentation
Presentation is loading. Please wait.
1
Non-Market Valuation Techniques
Revealed preference techniques - travel cost, hedonic pricing, cost of replacement Expressed preference techniques - traditional content valuation methods - discrete choice contingent valuation methods
2
Readings: Mitchell, R.C. and Carson, R.T. (1989) Using Surveys to Value Public Goods: The Contingent Valuation Method Resources for the Future, Washington, U.S. Garrod, G. and Willis, K.G. (1999) Economic Valuation of the Environment Edward Elgar, U.K. Chapters 5 and 6
3
Contingent Valuation Method
“The term contingent valuation is derived from the nature of the method: responses are sought from individuals as to their actions contingent on the occurrence of a particular hypothetical situation.” Gerrod and Willis (1999, 125.)
4
Non-use values Revealed preference techniques, such as travel cost or hedonic pricing, cannot value non-use values. This tends to be the domain of expressed preference techniques, such as contingent valuation.
5
Designing a CV Study When designing a CVM study consideration has to be given to the nature of the public good, payment method, and the context of valuation.
6
Six stages of conducting a CVM
1. Setting up the hypothetical market 2. Obtaining bids 3. Estimating mean WTP and/or WTAC 4. Estimating bid curves 5. Aggregating the data 6. Assessing the validity of the CV.
7
Stage One: Establishing the hypothetical market
Description of resource Reason for payment Bid vehicle – property tax, income tax or entry fees.
8
The good The CV should explicitly define the following aspects of the good: a. its attributes b. reference and target levels c. sources of the changes d. the extent and timing of change e. certainty of change
9
The payment method a. the payment vehicle – must be salient to the respondent – taxes, entry fees, rates. There is no point talking about increasing local council rates with tourists. b. the decision-making unit for which the value is to be expressed: hunting trip is an individual pursuit whereas a neighbourhood air quality is more a household issue. C. The timing of payments. d. The relevant prices of other goods.
10
The context of valuation
Participants are possibly influenced by who would benefit and who else would be offering their valuation. Whether the measure of value will be WTP or WTAC Knitch. Willingness to accept has been found to produce different answers to asking about willingness to pay.
11
Stage two: obtaining bids
Obtaining bids could be achieved by: A bidding game – higher and higher amounts are suggested to the respondents until their maximum WTP is reached. A payment card – respondents are presented with a range of values given income groups.
12
Stage two cont. Three different types of questioning Open ended
Iterative bidding Dichotomous choice
13
Open ended questions – Open-ended questions – respondents are asked for their willingness to pay. No value being suggested to them. Respondents have found it relative difficult to answer such questions, especially where they have no prior knowledge of trading with the commodity in question.
14
Suppose that the Brisbane City council was to increase rates to restore the wetlands of Moreton bay. How much would you be willing to pay?
15
Iterative bidding questions -
Suppose that the Brisbane City council was to increase rates to restore the wetlands of Moreton bay. How much would you be willing to pay? $10 per quarter If yes would you be willing to pay $20 per year If no would you be willing to pay $5 per year
16
Dichotomous choice questions:-
Would you be willing to pay an additional $10 per quarter in Brisbane City council rates to restore the wetlands of Moreton bay. Yes/ NO
17
Designing a CV Questionnaire:
6 Sections…
18
Sections Section One Section Two Section Three
Designed to familiarise the respondent with the situation, and policy issues. Section Two Ascertain those who knew and visited the area, extent of usage. Section Three Background information - explanatory information, maps, photo, etc.
19
Section Four Section Five Section Six
Payment vehicle - respondents must be presented with a believable payment vehicle and affected by. Section Five Ask the respondent for their willingness to pay. Section Six Collect socio-economic data
20
Don’t forget… Budgetary constraints
21
Sample Size The sample size is determined by the precision of sample statistics used as estimates of population parameters. The larger the sample, the greater the level of precision. The sample size is determined in part by the expected variance of the responses from pilot surveys. See Mitchell and Carson (1989) for a table of sample size requirements.
22
Stage three: estimating average WTP/WTAC
If open-ended, bidding game or payment card approach have been used, then the calculation of sample mean and/or median WTP or WTAC is straightforward.
23
It is usual in CVM to find that mean WTP exceeds median WTP, since the former is influenced by a relatively small number of relatively high bids (this is, the distribution of sample WTP is skewed). When dichotomous choice method has been used, then the calculation of average WTP is more difficult. The distribution of bids becomes binomial and depends of the shape of individual utility functions.
24
Stage four: estimating bid curves
WTPi = f(Yi, Ei, Ai, Qi) Income, Education, Age, environmental quality.
25
Stage five: aggregation of data
Aaggregation refers to the process whereby the mean bid or bids are converted to a population total value figure.
26
Issues: Choice of relevant population viz. All those whose utility will be significantly affected by the action Calculation of population mean – simple mean -. when the sample mean may be bias as a result of sample characteristics such as income or education, then the mean is estimated from the bid function by inserting what are considered appropriate population, income or education figures
27
. Time period over which the data should be aggregated. -
WTP, not WTAC, measures should be sought. Test should be made of whether WTP is sensitive to the level of environmental damage. Respondents should be reminded of their budget constraints. When the value of environmental benefits are over a given time frame then the benefits are discounted appropriately. Where irreversible environmental loss is involved, then the present value is calculated by taking a perpetuity (that is, by dividing the constant real benefit flow by the discount rate).
28
Descriptive Statistics
means/medians frequency tables bias regression logit regression
29
Problems with CVM Problems with CVM include: .biases .embedding .WTP/WTAC differences .Information effects .Transfer of benefit estimates
30
Biases There are three types of bias which may distort the results of a CV analysis: Design Bias Strategic Bias Starting point bias
31
Strategic bias respondents may think that they can influence the course of real events by the shape of their answers, and bias them accordingly.
32
Design bias The way in which the information about the problem is put across can be a powerful effect on the respondents (see Samples, Dixon, and Gowen, 1986, for a discussion of this problem in the context of endangered wildlife)
33
Example: Goto example This will need to be printed prior to the lecture. CV design bias.doc Divide the class into two groups. Present to one group the following: The government has discovered a threat of a major outbreak of aids that is expected to kill 600 people. Two programs have been proposed. Program A. If program A is adopted, 200 people will be saved. Program B. If program B is adopted there is a 1/3 probability that 600 people will be saved, and 2/3 probability that no one will be saved. And ask them to chose A or B. The result should be 70% for A and 30% for B. Present to group 2: Program C. If program A is adopted, 400 people will die. Program B. If program B is adopted there is a 1/3 probability that no one will die, and 2/3 probability that 600 people will die. The results should be the opposite dispite the fact that they are effectively the same. The difference being that one is termed in death while the other in life. See Gerrard and Willis p.156.
34
Example.. The government has discovered a threat of a major outbreak of a new aids virus that is expected to kill 600 thousand people. Two programs have been proposed. Program A. If program A is adopted, 200 thousand people will be saved. Program B. If program B is adopted there is a 1/3 probability that 600 thousand people will be saved, and 2/3 probability that no one will be saved. Choose A or B. Program C. If program C is adopted, 400 thousand people will die. Program D. If program D is adopted there is a 1/3 probability that no one will die, and 2/3 probability that 600 thousand people will die. Choose C or D.
35
Design bias cont. Expected Utility theory –
Principle of invariance that different representation of the same choice problem should produce the same preference set.
36
Starting point bias comes from the researcher suggesting starting bids.
37
Information There is no exogenous criterion to specify how much or in what context to provide information. Key: is the information leading or biased.
38
Validity of CVM “The validity of a measure is the degree to which it measures the theoretical construct under investigation. This construct is, in the nature of things, unobservable; all we can do is to obtain imperfect measures of that entity. In the contingent valuation context the theoretical construct is the maximum amount of money the respondents would actually pay for the public good if the appropriate market for the public good existed” (Mitchell and Carson, 1989, p. 190).
39
Content validity A measurement instrument has content validity if it accurately measures the aspects of the theoretical construct that are to be quantified. In the case of CVM, evaluating the content validity of a study involves examining the content of the survey instrument and related materials, such as visual aids to be sent to the respondent.
40
Does the description of the good and how it is to be paid for appear to be unambiguous?
Is it likely to be meaningful to the respondent? Is there anything in the scenario that might suggest to some respondents that the good would not be paid for?
41
Are the property rights and the market for the good defined in such a way that the respondents will accept the WTP format as plausible? Does the Scenario appear to force reluctant respondents to come up with WTP amounts?
42
Discrete Choice Methods
43
National Oceanic and Atmospheric Administration (NOAA)
A dichotomous choice format should be used. A minimum response rate from the target sample of 70 % should be achieved. in-person interviews should be employed Conservative design Elicitation format – WTP
44
Accurate description of the program or policy
Pre-testing of photographs Reminder of substitute commodities Adequate time lapse from accident – scenario of complete restoration must be viable.
45
Temporal averaging – select samples through time to account for time trends
“no Answer” option – A ‘no-answer” option should be explicitly allowed in addition to the “yes” and “no” options Follow up questions – to determine why they answered yes or no. Cross-tabulations – The survey should include a variety of other questions that help to interpret the responses to the primary valuation question.
46
account for … Income Prior knowledge of the site
Prior interest in the site Attitudes towards the environment Belief in the scenarios.
47
Basic Data Analysis Descriptive Statistics
Means, frequency tables and cross tabulations.
48
Advanced Data Analysis
The basic model for analyzing dichotomous CV responses is the Random Utility Model. Uij = ui(yj,zj,eij) Where I=1 is the state or condition that prevails when the CV program is implemented.
49
Random Utility Model Based on this model, respondent j answers yes to a required payment of tj if the utility with the CV program, net of the required payment, exceeds the utility of the status quo. U1(yj-tj,zj,e1j) >Uo(yj,zj,eoj) Pr(yesj) = pr(U1(yj-tj,zj,e1j) >Uo(yj,zj,eoj))
50
Reference Loomis J., Kent P., Strange L., Faush K. and A. Covich (2000). Measuring the total economic value of restoring ecosystem services in an impaired river basin: results from a contingent valuation survey. Ecological Economics 33(1):
51
Loomis et al. Ecosystem Services of a Plains River
“Five ecosystem services that could be restored along a 45-mile section of the Plate river..” Natural purification of water Erosion control Habitat for fish and wildlife Dilution of wastewater Recreational use
55
Measuring economic value
If the improvement in water quality.. Occurs in an urban environment – hedonic pricing models Primarily results in recreation benefits – Travel cost method – from which a demand curve and consumer surplus can be estimated. Where it results in existence and option values – Contingent valuation method
56
Survey design “If the South Platte Restoration Fund was on the ballot in the next election, and it cost your household $-- each month in a higher water bill would you vote in favor or against?” The $-- was filled with the following amounts 1,2,3,4,8,10,12,20,30,40,50,100
57
Model [log(yes)/(1-yes)] = bo – b1bid – b2 unlimited water
+ b3 government purchase +b4environmentalist -b5average water bill +b6urban
58
Interpretation Logit Regression Analysis of the Probability of Payment
Variable Co-efficient t-statistic Mean Constant 2.483 1.48 1 Bid Amount -0.144 -4.32*** 14.79 Unlimited water (corrected) 2.01** 0.452 Government Purchase 1.846 2.46** 0.78 Environmentalist 3.383 2.868*** Average water bill -0.063 -2.05** 35.80 Urban 1.803 2.55** McFadden R2 0.45 ** significant at the 0.05 level. Loomis (2000,113) *** significant at the 0.01 level.
59
Statistical Analysis - WTP
Probability (yes) = 1-(1+exp[Bo-B1($X)])-1 Mean WTP = (1/B1)*ln(1+ebo) Where B1 is the coefficient estimate on the bid amount Bo is either the estimated constant (where there are no other dependent variables) Or the sum of the estimated constant plus the product of the other independent variables and their means. Loomis (2000),
60
Calculation of WTP Mean WTP = (1/0.144)*ln(1+e2.91) = $21.05
61
Willingness to pay “… households would be willing to pay an average of $21 per month or $252 annually for the additional ecosystem services”
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.