CONTINGENT VALUATION METHOD:

BACK TO OUTLINE

As for all of the "Guide to Readings" pages, this list of questions is preliminary and incomplete. More readily available journals have been accessed first. This list will be updated.

Boxall, Peter C. et al., A Comparison of Stated Preference Methods for Environmental Valuation, Ecological Economics; 18(3), September 1996, 243- 53.

1. CV is one of several types of SP approaches, most of which have not been used widely in environmental valuation. Where have they been used?
2. In what paper was the choice theoretic interpretation of SP models formalized?
3. How does the research described in this paper differ from the earlier SP/RP paper in JEEM by Adamowicz et. al (1994)?
4. What three types of SP methods are distinguished in this paper? How are some common CV methods similar to these methods?
5. How are SP methods more general than CV methods?
6. How is the concept of "experimental design" used in SP applications?
7. These researchers find CV elfare estimates that are over twenty times higher than the alternative SP experimental choice estimate. What four reasons are given for why this result should be puzzling? What three possibilities are proposed for why this disparity is present in these data?
8. Which main factor do these authors conclude probably accounts for the extremely high CV welfare estimates?
9. Why do these authors argue that SP methods should be more widely used for nonmarket valuation?

Hanley, Nick; Milne, Jennifer, Ethical Beliefs and Behaviour in Contingent Valuation Surveys, Journal of Environmental Planning and Management; 39(2), June 1996, 255-72.

1. This paper introduces the thorny problem (for conventional economics) of the possibility of lexicographic or rights-based views. What are these views?
2. In what three ways is the environment generally modelled as entering the consumer's utility function?
3. How is the utilitarian perspective fundamental to stanrd welfare analysis?
4. What is the implication of rights-based beliefs for individual's willingness to trade dollars for environmental changes?
5. How do these researchers look for evidence of rights-based views? Do they appear to exist?
6. Do these authors propose any concrete model allowing for rights-based views that generalizes the usual Kaldor-Hicks criterion on which cost-benefit analysis is based?

Roe, Brian; Boyle, Kevin J.; Teisl, Mario F., "Using Conjoint Analysis to Derive Estimates of Compensating Variation," Journal of Environmental Economics and Management; 31(2), September 1996, 145-59.

1. In the terminology of conjoint analysis (the marketing term for what environmental economists have begun calling stated preference or stated choice methods), what is a "part-worth"?
2. Have analyses of conjoint ratings data typically used utility-theoretic empirical specifications?
3. How do standard conjoint ratings studies estimate their models?
4. What is a double-hurdle Tobit model?
5. How do these authors convert conjoint ratings data into a form amenable to analysis in a standard random utility framework?
6. What is the concern about "centering"?
7. Were separate samples used to estimate welfare measures by ratings, ranking, and binary methods? Explain.
8. What do these authors recommend concerning the future of conjoint analysis techniques for non-market resource valuation?

Smith, V. Kerry; Osborne, Laura L., Do Contingent Valuation Estimates Pass a "Scope" Test? A Meta analysis, Journal of Environmental Economics and Management; 31(3), November 1996, 287-301.

1. What is "internal consistency" as it concerns contingent valuation methods?
2. Was new survey research conducted for this study?
3. What is a meta-analysis?
4. How precise was the NOAA Blue Ribbon Panel on the issue of scope-test requirements for CV?
5. Which specific environmental good is the focus of this study?
6. Briefly describe the strategy used to conduct this very formal quantitative literature review.

Herriges, Joseph A.; Shogren, Jason F., "Starting Point Bias in Dichotomous Choice Valuation with Follow-Up Questioning," Journal of Environmental Economics and Management, 30(1), January 1996, pages 112-31.

1. Distinguish between single- and double-bounded dichotomous choice contingent valuation formats.
2. What disadvantage of single-bounded formats makes double-bounded formats attractive?
3. Double-bounded formats are a limited form of what other approach to valuation elicitation?
4. What is starting point bias? Why might it affect respondents who are uncertain about the value of the good?
5. What environmental good is valued in their application, and what type of survey (in-person, mail or phone) was used?
6. What "treatments" were employed in order to allow examination of starting point bias?
7. What is the distinction between "design" bids and "effective" bids, with anchoring on the starting point?
8. What do the Monte Carlo simulations suggest about bias in estimated WTP in "one-way street up", "one-way street down", and double-bounded formats? Upon what assumptions does the finding wrt double-bounded formats depend?
9. What is the effect on estimated dispersion of WTP?
10. What are the implications for efficiency gains from follow-up questions?
11. How do the authors propose that respondents may update their uncertain values in response to incoming bids?
12. What advice do the authors have concerning future research in this area?

Smith, V. Kerry, "Can Contingent Valuation Distinguish Economic Values for Different Public Goods?," Land Economics, 72(2), May 1996, pages 139-51.

1. Which of three characteristics of acceptable CV estimates does the author emphasize in this paper?
2. Why is this issue relevant to concerns about CV responses simply reflecting willingness to contribute to a "good cause"?
3. Why does the earlier literature not allow any clearcut assessment of whether CV respondents can discriminate between important and unimportant environmental goods?
4. What does Smith identify as two factors that describe the elements of a choice?
5. What two environmental goods are offered in this study?
6. What type of survey was employed? What sampling method was used? What payment vehicle was suggested?
7. Were respondents able to discriminate? What three types of statistical tests were used to make this assessment?
8. How does sensitivity to (a) the size of the registration surcharge and (b) the number of vehicles a respondent had registered imply a rejection of the "contributions framework"?

Harrison, Glenn W.; Lesley, James C., "Must Contingent Valuation Surveys Cost So Much?," Journal of Environmental Economics and Management, 31(1), July 1996, 79-95.

1. About how much was the reported cost of the contingent valuation study undertaken for the state of Alaska in support of the plaintiffs in the Exxon Valdez case?
2. About how much was the study's "conservative" estimate of the economic damages from the oil spill?
3. Why do the authors argue that it is important to figure out how to save money when doing CV studies (two reasons)?
4. What sort of valuation function do the authors advocate if one's objective is to simulate population WTP from a convenience sample?
5. What about multicollinearity and individually insignificant slope coefficients in this context?
6. What kind of convenience sample was used for this paper?
7. How well does the convenience-sample WTP function predict the expensive representative-sample Exxon Valdez results when applied to the population demographic distribution?
8. Do these results generalize to all "gold-plated" CV studies?

Carson, Richard T. et al., "Contingent Valuation and Revealed Preference Methodologies: Comparing the Estimates for Quasi-Public Goods," Land Economics, 72(1), February 1996, 80-99.

1. How is this type of research described?
2. From how big a pool of CV studies were these comparison studies drawn?
3. How many comparisons of CV and revealed preference non-market value estimates are employed in this study?
4. Distinguish between "convergent validity" and "criterion validity."
5. What types of studies were excluded from comparison?
6. What five broad categories of revealed preference analysis are involved?
7. What three classes of quasi-public goods are identified?
8. What "variable" is analyzed, with observations being individual comparisons of methods?
9. What does the distribution of this variable look like, roughly?
10. Does this variable depend of the type of RP method or the nature of the good being valued, on time, or on the publication status of the study?
11. Roughly how highly correlated are the CV and RP estimates across all these comparisons?
12. What do these results imply about the policy proposition that any CV estimate should be halved before being used in decision-making?

Ready, Richard C.; Buzby, Jean C.; Hu, Dayuan, "Differences between Continuous and Discrete Contingent Value Estimates," Land Economics, 72(3), August 1996, 397-411.

1. What is the argument for using "split-samples" when comparing open-ended and closed-ended elicitation methods?
2. In this study, how were item non-response rates related to bid amounts in the discrete choice CV instrument? What would be the implications of non-response bias in this case?
3. What is "yea-saying" and how might it bias DC estimates?
4. How might free-riding influence open-ended responses?
5. What application is used in this study, and what types of elicitation methods are used?
6. How are "synthetic" DC data constructed?
7. Summarize the results; are payment card and dichotomous choice results identical? are scope effects apparent? is yea-saying apparent? What conclusion is drawn about the prefernce for DC methods over continuous elicitation methods?

Li, Chuan Zhong; Mattsson, Leif, "Discrete Choice under Preference Uncertainty: An Improved Structural Model for Contingent Valuation," Journal of Environmental Economics and Management, 28(2), March 1995, 256-69.

1. How are the theoretical frameworks of Hanemann and Cameron generalized in this paper?
2. How is each respondent's confidence in their discrete choice response measured?
3. What is the role of the "equivalent deviate" when respondent's subjective confidence in their answers is incorporated into the analysis?
4. What is estimated in the modified model? Using what software?

Ready, Richard C.; Whitehead, John C.; Blomquist, Glenn C., "Contingent Valuation When Respondents Are Ambivalent," Journal of Environmental Economics and Management, 29(2), September 1995, 181-96.

1. In most dichotomous choice CV contexts, respondents start from baseline environmental quality and baseline income, and are typically asked their WTP for higher environmental quality at the cost of lower income. What sort of a move does this represent in money-amenity space?
2. How is an "ambivalence region" related to indifference curves?
3. What possible decision rules might be adopted by ambivalent respondents?
4. How did the treatments differ across the two split samples in the application used for this study?
5. Do question formats (dichotomous versus polychotomous) have any effect on response rates?
6. Why were ordered logits (or probits) not used in this study? Would they not be more appropriate?
7. What would "conservatism" suggest for the point in the ambivalence region where respondents switch from yes to no when forced to make a choice?
8. What do the study results imply for the NOAA Panel recommendation that a "no-vote" option be included? How did respondents seem to react to the opportunity "not to think hard" about their values?
9. How do the study's conclusions suggest that dichotomous choice CV surveys could be improved?

Ready, Richard C.; Hu, Dayuan, "Statistical Approaches to the Fat Tail Problem for Dichotomous Choice Contingent Valuation," Land Economics, 71(4), November 1995, 491-99.

1. Why are "fat tails" a problem in dichotomous choice CV studies?
2. What truncation strategies have been advocated, and what are their consequences for estimated CV WTP values?
3. What "normalization" of this truncation has been proposed, but what is its disadvantage?
4. What is a "pinched logit"? What are its advantages over a non- parametric approach?
5. What do the authors recommend in the event of insufficient data at high bid levels (a common problem)?

Holmes, Thomas P.; Kramer, Randall A., "An Independent Sample Test of Yea-Saying and Starting Point Bias in Dichotomous-Choice Contingent Valuation," Journal of Environmental Economics and Management, 29(1), July 1995, 121-32.

1. What is "procedural variance"?
2. How are bootstrap techniques used in the Monte Carlo simulations in this paper?
3. What issue does this procedure allow the researchers to get at?
4. Why are sample selection corrections necessary to evaluate yea-saying?

Loomis, John; Gonzalez Caban, Armando; Gregory, Robin, "Do Reminders of Substitutes and Budget Constraints Influence Contingent Valuation Estimates?," Land Economics, 70(4), November 1994, 499-506. plus Whitehead, John C.; Blomquist, Glenn C., "Do Reminders of Substitutes and Budget Constraints Influence Contingent Valuation Estimates? Comment," Land Economics, 71(4), November 1995, 541-43. plus Loomis, John; Gonzalez Caban, Armando; Gregory, Robin, "Do Reminders of Substitutes and Budget Constraints Influence Contingent Valuation Estimates? Reply," Land Economics, 71(4), November 1995, 544-45.

1. What was the motivation for this study?
2. How were survey instruments designed in order to test this issue? What was the application?
3. How were the response rates? Might they be expected to differ across treatments? Was this assessed or compensated-for?
4. What are the findings of the original study?
5. What prompted the comment by Whitehead and Blomquist?
6. What were the results from Whitehead and Blomquist's replication of their earlier study?
7. In the reply, what is discussed concerning the nature of "other goods" brought to the attention of respondents before eliciting values?

Whitehead, John C., "Willingness to Pay for Quality Improvements: Comparative Statics and Interpretation of Contingent Valuation Results," Land Economics, 71(2), May 1995, 207-15.

1. Ted McConnell, in a 1990 paper, demonstrated that if the marginal utility of income is constant, the Hanemann and Cameron approaches to dichotomous choice contingent valuation are linear transforms of each other. What generalization of McConnell's analysis is offered in this paper?
2. How are users and nonusers distinguished in this analysis?
3. Inventory the comparative statics results.
4. What are the implications for empirical applications employing sufficient variables to conform with the theoretical exposition?
5. How do the results of the derivations suggest theoretical validity tests?
6. How do they suggest tests for convergent validity with other recreational demand models?

Gregory, Robin et al., "How Precise Are Monetary Representations of Environmental Improvements?," Land Economics, 71(4), November 1995, 462-73.

1. In this study, respondents are asked to value environmental goods in multiples or fractions of an arbitrarily assigned budgetary unit (e.g. $1 or $20). Were people universally able to do the arithmetic?
2. Are there any potentally interesting sample selection consequences from excluding people who cannot do the arithmetic?
3. Was the distribution of implied values stable across groups who where asked to work in terms of different "budgetary units"?
4. What are some possible explanations for the findings in this study?
5. How precise are the implications for contingent valuation survey methods?

Elnagheeb, Abdelmoneim H.; Jordan, Jeffrey L., "Comparing Three Approaches That Generate Bids for the Referendum Contingent Valuation Method," Journal of Environmental Economics and Management, 29(1), July 1995, 92-104.

1. What phase of contingent valuation survey design is addressed in this study?
2. What is the Boyle et al. approach to generating bids?
3. What is the Cooper approach to generating bids?
4. What additional technique does this paper suggest?
5. Outline how Monte Carlo techniques are used to assess these different methods.
6. Under the controlled conditions in these experiments, which method produced the least biased results? Which was most efficient?
7. Is careful processing of pretest valuation data via the Boyle et al. or Cooper methods preferable to the ad hoc choice and assignment of bids?
8. Which of the two methods is easier to implement and when can it best be used?

Kanninen, Barbara J., "Bias in Discrete Response Contingent Valuation," Journal of Environmental Economics and Management, 28(1), January 1995, 114-25.

1. What aggravating property of discrete-choice contingent valuation studies is addressed in this paper?
2. Is bias due to bid design typically large? Is it systematic?
3. Can bias due to bid design be properly assessed across a set of actual studies? By Monte Carlo methods? Analytically?
4. What limits on bids have been suggested in prior research?
5. How does one test for outliers in discrete choice models?
6. What is the best way to avoid bias and large variance in discrete choice models?
7. What rule-of-thumb prescription does the paper offer, and is this the same as "optimal bid design"?

Carson, Richard T.; Mitchell, Robert Cameron, "Sequencing and Nesting in Contingent Valuation Surveys," Journal of Environmental Economics and Management, 28(2), March 1995, 155-73.

1. What vaguely defined property of contingent valuation studies do the authors attempt to clarify and investigate in this paper?
2. How do the authors define "sequencing" and "nesting"?
3. Explain four types of amenity misspecification errors that might lead to an appearance of embedding problems in empirical CV research.
4. How do the authors interpret the embedding that appears to exist in the Desvousges et al. "2000, 10,000, and 200,000 bird" study?
5. What criteria do the authors advocate for CV "component sensitivity" tests?

Teisl, Mario F. et al. "Test-Test Reliability of Contingent Valuation with Independent Sample Pretest and Posttest Control Groups," American Journal of Agricultural Economics, 77(3), August 1995, 613-19.

1. What concern about contingent valuation estimates of environmental values is addressed in this paper?
2. What methodology is used to assess this potential problem?
3. What are the findings? Do they hold for both users and non-users of the resource?
4. Is the conclusion of reliability an artifact of respondent recall, or simply a statistical anomaly? Why or why not?