Download - Contingent Valuation in a Policy Context
-
7/31/2019 Contingent Valuation in a Policy Context
1/31
Click to edit Master subtitle style
5/11/12
Cont ngent Va uat on n a Po cyContext:
The National Oceanic and
Atmospheric AdministrationReport and Its Implications for
The Use of Contingent Valuation
Methods in Policy Analysis inBritainKen Willis
University of Newcastle upon TynePresented by.
Setia Lesmana H351100034Yocie Gusman H351100044Baidhuri Purna Edi H351100064
-
7/31/2019 Contingent Valuation in a Policy Context
2/31
Click to edit Master subtitle style
5/11/12
23 Maret
-
7/31/2019 Contingent Valuation in a Policy Context
3/31
Click to edit Master subtitle style
5/11/12
-
7/31/2019 Contingent Valuation in a Policy Context
4/31
Click to edit Master subtitle style
5/11/12
-
7/31/2019 Contingent Valuation in a Policy Context
5/31
Click to edit Master subtitle style
5/11/12
-
7/31/2019 Contingent Valuation in a Policy Context
6/31
5/11/12
Outline: Introduction
Detailed Conclusions of The NOAA Report
General guide-lines
Value elicitation surveys
Goal for value elicitation surveys
Recommendations Currently in Use
Recommendations Not Generally in Current
Use Referendum versus Alternative Elicitation
Methods
Embedding Problems
Warm Glow Effects
-
7/31/2019 Contingent Valuation in a Policy Context
7/31
5/11/12
Introduction
ExxonValdez: oilspilled(24-03-1989)
Losses of marketed goods, e.g. tofishermen;
Restoration of natural resourcesystem&loss of non use values
CERCLA 1980 => sued fordamaged use&non use values byCVMOil Pollution Act 1990 =>procedures for assessing damages
NOAA commisioned a blue ribbonpanel :Use&non use valuesNOAA report 1993 => qualifiedrecognition CVM +
recommendations
-
7/31/2019 Contingent Valuation in a Policy Context
8/31
5/11/12
Detailed Conclusions of The NOAAReport
Principal Conclusion:
CV studies convey usefulinformation as reliable bystandards that seem to beimplicit in similar contexts, like
market analysis for new andinnovative products, and theassesssment of other damagesnormally allowed
The appropriate... Agencies shouldbegin to accumulate standarddamage assessments for a rangeoil spills... That process shouldimprove the reliability of CV
studies in damage assessment. It
Recommendation General
guide-lines Value
elicitationsurveys
Goals forvalueelicitationsurveys
-
7/31/2019 Contingent Valuation in a Policy Context
9/31
5/11/12
General guide-lines
1. Sample type and size: that
probability sampling is essential.2. Non-response:this should be
minimized otherwise survey results
will be unreliable.3. Personal interview:it is unlikely
that reliable estimates of values can
be elicited with mail surveys. Face-to-face surveys are preferable,although telephone interviews havesome advantages in terms of costand centralized supervision.
-
7/31/2019 Contingent Valuation in a Policy Context
10/31
5/11/12
Value elicitation surveys
1. Conservative design: increasesthe realibility by eliminatingextreme response that can enlargeestimated values. Thus, an option
which tends to underestimatewillingness to pay(WTP) ispreferred.
2. Elicitation format: WTP should beused instead ofwillingness to accept(WTA), because the former is the
conservative choice.
-
7/31/2019 Contingent Valuation in a Policy Context
11/31
5/11/12
Value elicitation surveys7. Adequate time lapse from accident: to avoid
misunderstanding of restoration possibilities, andrespondents reporting a substantial passive use losseven when informed full restoration will occur.Questionnaire should force respondents to considerthe difference between interim and steady statepassive use value.
8. Temporal averaging: to reduce measurement noise.A time trend in responses would cast doubt upon thereliability of the findings.
9. No-answer option: to allow for approximate
indifference, inability to answer without moreinformation, preference for another mechanism, andboredom with the survey.
10. Yes/no follow ups: to ascertain why respondentsanswered yes or no to a WTP question.
11. Cross tabulations: to interpret WTP responses in
-
7/31/2019 Contingent Valuation in a Policy Context
12/31
5/11/12
Goal for value elicitation surveys
1. Alternative expenditure possibilities:
respondents should be reminded their WTP forthe environmental good would reduce theirexpenditure on other private goods.
2. Deflection of transaction value: the survey
should be designed to deflect warm-glow effect.Utility derived from charitable giving may comemainly from the act of giving rather than thematerial change that follow the gift. While both
are real values, there may be close substitutesto cleaning up oil spills which would producethe same charitable warm-glow efffects.
3. Steady state or interim losses: respondents
should be able to distinguish between these,
-
7/31/2019 Contingent Valuation in a Policy Context
13/31
5/11/12
Goal for value elicitation surveys
4. Present value calculations of interim
losses: it should be demonstrated that therespondents are sensitive to the timing ofrestoration
5. Advance approval: the CV survey should be
approved by both sides in the legal action.
6. Burden of proof: to rest with CV designers, todemonstrate that the CV survey is reliable.
7. Reliable reference surveys: governmentshould create reliable reference surveys tointerpret the Panels guide-lines and calibratesurveys in meeting their conditions
-
7/31/2019 Contingent Valuation in a Policy Context
14/31
5/11/12
Recommendations Currently in Use
No Recommendation
s
Remarks
1 Personalinterview
*(Macmillan et al.,1994)=>mailsurveys
2 Careful
pretesting forCV questinnaire
Pretested in pilot surveys
3 Elicitationformat
Prefer WTP to WTA>Divergence between WTP &WTA (Bateman and Turner, 1993)>Explanation: questionnairedesign&interviewing techniques;respondents rejection of the
property right; prospect theory
-
7/31/2019 Contingent Valuation in a Policy Context
15/31
5/11/12
Recommendations Currently in Use
No Recommendation
s
Remarks
4 Accuratedescription ofprogramme or
policy
Using brochure (Ecotec, 1993;Loomis and du Vair, 1993; Williset al., 1993a)
5 Pretesting ofphotographs
e.g. Photograph of the SomersetLevel and MoorsEnvironmentally Sensitive Area(ESA) (Willis et al., 1993a);montages of coastal erosion(Tunstall and Cooker,1992)
-
7/31/2019 Contingent Valuation in a Policy Context
16/31
5/11/12
Recommendations Currently in UseNo Recommendatio
nsRemarks
6 Yes/no followups
e.g. Studies of flood protection in thebroads (Bateman et al., 1992,1993);landscape, wildlife and historicalarchaeological preservation benefits oflandscapes through ESA prescription
(Willis et al., 1993a), benefits low flowalleviation in the River Darent (Willisand Garrod, 1993c), the protection ofthe aquatic environment from acid rain(Ecotec, 1993)
7 Crosstabulations
*To asses the relationship of WTP withexplanatory variable => modelling.*LF => r2 => extremely low (Cobbingand Slee, 1993)-> the Mar Lodge Estate*Good questionnaire&interviewer
training (Willis et al., 1993a)*Dichotomous choice models => hi h
-
7/31/2019 Contingent Valuation in a Policy Context
17/31
5/11/12
Recommendations Not Generally inCurrent Use
No Recommendation
s
Remarks
1 Sample typeand size =>large sampledsize (around1000)
Cummings et al. (1986): a smallsample size (160 or so),OE,levelof statistical realiability => lieWTP 20%, true WTP 80%*NOAA Panel: no consideration=> complexities samplingstrategies
2 ReferendumFormat
*Bateman et al. (1992) => theNorfolk Broads (3000 sample;split>> OE-iterative biding-Dichotomous Choice)
*Willis et al. (1993a) => ESAs(3000 sample)
ecommen a ons o enera y n
-
7/31/2019 Contingent Valuation in a Policy Context
18/31
5/11/12
ecommen a ons o enera y nCurrent Use
No Recommendations
Remarks
3 Pretesting forinterviewereffects =>modify the
standard face-to-face survey toallowrespondents to
either: (i) writetheir vote on aballot anddeposit it in asealed box; or (2)
mail their ballots
*a large pilot survey; impractical*in the main survey: modernstatistical packages=>SAS
ecommen a ons o enera y n
-
7/31/2019 Contingent Valuation in a Policy Context
19/31
5/11/12
ecommen a ons o enera y nCurrent UseNo Recommendation
sRemarks
4 Reporting *questionnaires available toother researchers*data sets=> original researcher*not open to subsequentscrunity&interpretation
5 Temporalaveraging
*at different points in time =>not undertaken*creates immediate emotionalconcern, and/or losses are likelyto be interim; i.e. Theenvironment will recoverovertime => undertaken
6 Reminder of
undamaged
*Willis and Garrod (1993c) =>
low flow allevation in rivers*
-
7/31/2019 Contingent Valuation in a Policy Context
20/31
5/11/12
Rare agreement:
questionnaire by 2 opposing partieshaving interest (producing a
favourable outcome)
-
7/31/2019 Contingent Valuation in a Policy Context
21/31
5/11/12
Referendum vs AlternativeElicitation Methods
Mitchell and Carson(1989)
-
7/31/2019 Contingent Valuation in a Policy Context
22/31
5/11/12
Referendum vs AlternativeElicitation Methods
OE => unlikely provide the mostreliable valuation for non-use(caused by: scenario lack realism;strategic overstatement)
Mitchell and Carson (1989): strategicbias (need: analysis of outlier, testsfor bimodal distribution; tests of sub-
samples) Free-riding => underestimation of
WTP (minimized by introducing the
risk of potential exclusion from the
-
7/31/2019 Contingent Valuation in a Policy Context
23/31
5/11/12
Embedding Problems
Observed firstly by Kahneman andKnetsch (1984): to clean up lakes inMuskoka & all lake in Ontario.
Produced by careless questionnaiedesign.
Reflected that different levels ofprovision of the good had not beenclearly specified to respondents.
Suggestion: increasinginformation&context.
-
7/31/2019 Contingent Valuation in a Policy Context
24/31
5/11/12
Embedding Problems
How much information/context: noexogenous criterion.
Hoevenagel and van der Linden (1993):significantly different WTP values (as
economic theory predicts) CV responses cannot be context free
=> the size & nature of the choice
set. Randall (1991):
market prices: conditional (depend on
institutions, supply-demand conditions andex ectations about both .
-
7/31/2019 Contingent Valuation in a Policy Context
25/31
5/11/12
Warm Glow Effects
Panels suggestion: people onsupport one/two charitableorganizations to the tune $10 to $20
per year; 99,9% charities =>nosupport from any individual. CVM: forminor public goods or a public good
with large numbers of substitutes =>zero WTP.
Donation =>> WTP to avoid
guilt/embarrasment or to buy warm
-
7/31/2019 Contingent Valuation in a Policy Context
26/31
5/11/12
Some Framing Issues
Valuation question should be framedin terms of WTP, not WTA (no logicalreason)
Precautionary principle (methods ofrisk assessment) => Risk ofirreversible or catasthrophic
environmental effects (ex. Safeminimum standard)
Tversky and Kahneman (1982): in terms
of lives saved or in terms of lives lost
-
7/31/2019 Contingent Valuation in a Policy Context
27/31
5/11/12
Some Framing Issues
Embedding problems
MAFF (1992) -> value benefit 2 ESAs:the South Downs and Somerset Level
and Moors (perfect no substitute foranother)
Sites of Special Scientific Interest (SSSIs)
or low flow allevation (LFA) in rivers(perfect subtitute for each other)
Modify national resouce accounts by
applying values of environmental
-
7/31/2019 Contingent Valuation in a Policy Context
28/31
5/11/12
Other Issues
The items of concern to psychologists,economists, statisticans.
Psychologists:
1. Covariation misestimation
2. Perceptions of environmental
hazards3. Lack of awareness
4. The effect of information on CV
estimates, in terms of:
-
7/31/2019 Contingent Valuation in a Policy Context
29/31
5/11/12
Other Issues
Economists:
1) Paradigm in which people areassumed to be able to articulate andexpress values.
The Concerns: strategic response, etc
Treatment: proper incentives,referendum models, etc.
Theoretical base: demand analysis, etc.
Tests of success: sensible answer
-
7/31/2019 Contingent Valuation in a Policy Context
30/31
5/11/12
CV and the Law
Environmental damage in Englishlaw: criminal law (e.g. under theWildlife and Countryside Act 1981,
and other Act) & tort law. Compensation -> loss of value -> the
market value of the lost
resources+transaction cost. Non-pecuniary losses (e.g. in
relation to health effects) ->
arbitrary.
-
7/31/2019 Contingent Valuation in a Policy Context
31/31