Abstract: Follow-up questions are frequently used to improve the efficiency of dichotomous choice contingent valuation questionnaires. However, a number of authors have noted a significant difference between the WTP distributions implied by initial and follow-up question responses. This paper investigates starting point bias as one explanation for the phenomenon. We develop a model of starting point bias in the context of dichotomous choice questioning with follow-up, using a Monte Carlo simulation to illustrate the potential bias imparted to WTP estimates. An application is provided based on data from a CV study in northcentral Iowa.

Abstract: This paper reports the first test evaluating whether stated choices as part of a contingent valuation survey discriminate between significant and trivial causes. Using a random digit dialed sample for North Carolina, two different plans were posed to respondents--one to expand a popular North Carolina highway flower planting program nationwide and a second to facilitate the use of recycled tires in making asphalt for highways. Random assignment, identical payment mechanisms, and overall question designs were used. The results indicate clear discrimination in choices, choice functions, and the estimated willingness-to-pay for the two plans.

Abstract: A common assumption in the economic valuation of the environment is that environmental resources can essentially be treated identically to produced goods and services in estimating welfare measures. The most critical aspect of this assumption in this regard is that individuals are willing to trade-off environmental quality changes in exchange for changes in their income. In this paper, we briefly review some of the conceptual arguments over this issue, and existing empirical evidence, before presenting the results of a new pilot study which addresses some of the possible problems in viewing the environment in this fashion.

Abstract: Thirty-two previously published papers examine measures of the economic values for the services of nature and how those values are constructed from people's choices. Papers focus on overviews of economic theory and econometric methods of environmental valuation; travel cost recreation demand models; hedonic models; household production models; the contingent valuation method; and the agenda of future research on nonmarket valuation. Smith is the Arts and Sciences Professor of Environmental Economics at Duke University and a University Fellow with Resources for the Future.

Abstract: This article presents an experiment designed to test the reliability of the benefit function transfer approach using contingent valuation methods. The experiment uses data collected from anglers surveyed across eight contiguous Texas Gulf Coast bay regions over three distinct time periods. Results indicate that the benefit function transfer approach tends to over-estimate benefits, implying that, at least for the case of recreational saltwater fishing Texas, the benefit function transfer approach is not reliable.

Abstract: Construct validity is concerned with the consistency of empirical measures with theoretical relationships. This paper reviews and extends the theoretical relationship between averting costs and willingness-to-pay. Measures of these two theoretical constructs with empirical averting costs savings and contingent valuation from one population are used to establish construct validity of the averting cost technique used. Empirical results indicate that the contingent valuation and averting costs measures have a low correlation. Although low correlation is consistent with the construct validity of averting costs measures, these results indicate that inferences from averting cost measures are quite limited. Coauthors are Wesley N. Musser, James S. Shortle, and Lynn M. Musser.

Abstract: A literature search provides 83 studies from which 616 comparisons of contingent valuation (CV) to revealed preference (RP) estimates are made. Summary statistics of the CV/RP ratios are provided for the complete dataset, a 5 percent trimmed dataset, and a weighted dataset that gives equal weight to each study rather than each CV/RP comparison. For the complete dataset, the sample mean CV/RP ratio is 0.89 with a 95 percent confidence interval [0.81-0.96] and a median of 0.75. For the trimmed and weighted data sets, these summary statistics are (0.77; [0.74-0.81]; 0.75) and (0.92; [0.81-1.03]; 0.94), respectively. The Spearman rank correlation coefficients between the CV and RP estimates for the three datasets are 0.78, 0.88, and 0.92. Coauthors are Nicholas E. Flores, Kerry M. Martin, and Jennifer L. Wright.

Abstract: For an internal consistency check, one can examine WTP responses to different surveys. Holding an environmental problem constant, concave preferences imply WTP concave success of a remedy. The same preferences imply WTP convex in the magnitude of an environmental problem, assuming a successful remedy. Having opposite signed curvature of these two WTP patterns is general, not requiring concave preferences. For an adding-up test, an environmental commodity can be divided. Then, WTP for one part plus WTP for the second part, conditional on already having the first part, equals WTP for the whole, adjusted by an income effect.

Abstract: In a split-sample contingent valuation study of willingness to pay for food safety improvements, the dichotomous choice elicitation method consistently generated much larger estimates of willingness to pay than did a continuous method. Little or none of these differences was due to bias introduced by the statistical techniques used with the dichotomous choice data. Most or all of the difference was due to differences in respondent behavior. In addition, the continuous willingness to pay responses showed a significant scope effect, while the dichotomous choice responses did not. The observed difference in behavior may be attributable in part to yea-saying by dichotomous choice respondents.

Abstract: Twenty papers, originating at a conference held in Washington, D.C., in June 1993, assess the value consumers and society place on food safety and nutritional quality and evaluate the major methodologies being employed in valuation research. Papers focus on research perspectives and theoretical models, a comparison of valuation methodologies; a close examination of contingent valuation; and inputs to valuation studies. Contributors are mainly economists. Caswell is a professor in the Department of Resource Economics at the University of Massachusetts, Amherst.

Abstract: This paper is concerned with discrete choice contingent value estimation in the presence of response errors. We attempt to extend the theoretical frameworks developed by Hanemann (1984) and Cameron (1988) by introducing preference uncertainty with individual respondents. We assume that respondents have incomplete knowledge about their true valuation of a nonmarket resource and thus may give wrong yes/no answers on whether or not to pay a given bid amount for the resource. A post-decisional confidence measure for each observed yes/no answer is elicited for characterizing the degree of preference uncertainty. With the help of this confidence measure, we develop an improved structural model for deriving value estimates. Data from a forest environment valuation survey are used for the empirical estimations.

Abstract: Eight papers, resulting from a conference held at the University of Neuchatel in October 1994, examine the contingent valuation method as applied to road accidents. Nathalie G. Schwab Christe explores the valuation of human costs by the contingent method in Switzerland. Kristian Kidholm presents the Danish survey of the value of traffic safety. Ulf Persson, Anna Lugner Norinder, and Marianne Svensson focus on the Swedish experience of valuing the benefits of reducing the risk of non-fatal road injuries. Brigitte Desaigues and Ari Rabl discuss an econometric analysis of a contingent valuation in France. Michael W. Jones-Lee, Graham Loomes, and Angela Robinson question why two theoretically equivalent methods produced two very different values. W. Richard Dubourg assesses whether preferences for safety are too imprecise for contingent valuation. Michael W. Jones-Lee and Graham Loomes address preference-based values of safety for public transport modes. W. Kip Viscusi considers the automobile risk metric for valuing health risks. Schwab Christe and Soguel are with the Institute for Regional and Economic Research, University of Neuchatel.

Abstract: Respondents to contingent valuation surveys may have difficulty resolving ambivalence over trade-offs between money and changes in levels of environmental amenities. In two separate studies, respondents were given the opportunity to express intensity of preferences through polychotomous choice valuation questions. This elicitation method produced slightly higher rates of usable responses than dichotomous choice, but with wide ambivalence regions. Dichotomous choice responses appear to reflect a conservatism strategy, in that ambivalent respondents tend to reject any move away from the baseline.

Abstract: If a cumulative density function estimated from dichotomous choice data has an unrealistically fat right-hand tail, mean willingness to pay (WTP) will be overestimated. Truncating the range of integration results in a lower-bound estimate of the true mean WTP. A normalization procedure proposed by K. J. Boyle, M. P. Welsh, and R. C. Bishop (1988) unnecessarily depresses the estimate further. A new statistical approach is presented that allows a best guess estimate of mean WTP. Because this estimate involves extrapolation beyond the range of the data, it tends to be highly variable.

Abstract: Using an original data set, this article presents the first effort to compare systematically the values of a statistical life obtained from the two main methodologies existing in the literature - the revealed-preference approach (wage-risk analysis) and the contingent-valuation approach - applied to the same sample of individuals. A survey was conducted in the Montreal area with a questionnaire containing two series of questions: one set of contingent-valuation questions and one set of questions pertaining to the information necessary for performing a wage-risk study. The results show that the values of a statistical life obtained from each method are different from each other.

Abstract: This paper uses a referendum-style survey approach known as dichotomous-choice contingent valuation to estimate the benefits of restricting the uses of 6.9 million acres of desert land. Statistical techniques estimate the value to California residents of creating three new national parks and 76 new wildnerness areas in the high and low deserts of eastern California. The total amount that California residents would be willing to pay to enact desert protection legislation ranges from $177 million to $448 million per year. This estimate hinges on the assumptions that (1) the residents who did not complete and return the survey questionnaire ("nonrespondents") would receive no benefits from desert protection and (2) the estimate of willingness to pay for the "respondents" is unbiased.

Abstract: We developed several diagnostic tools to test the convergent validity of two common contingent valuation elicitation procedures. Data were collected from independent samples receiving dichotomous choice and payment card questions. We compared actual with counterfactual responses using deterministic and Monte Carlo methods and we found that WTP distributions and mean values varied by the value elicitation method. We also developed a paired-comparison test for procedure variance which indicated that yea-saying and starting point bias influenced dichotomous choice responses.

Abstract: This paper extends variation function theory by examining the effects of changes in prices, quality, and income on willingness-to-pay for quality change. Comparative static effects are found for both on-site users and nonusers of the resource. These results are used to interpret contingent valuation empirical models. For example, the substitution and complementary relationships between trips to natural resource sites can be identified. This paper also suggests tests for comparison of contingent valuation with recreation demand models and other tests of validity.

Abstract: The paper tests two alternative specifications for the opportunity cost of time in travel cost models. The standard travel cost survey design is enriched to include a contingent valuation type question about peoples' willingness to accept compensation to forgo a precisely defined recreational experience. It is hypothesized that individually revealed value of time more appropriately reflects the opportunity costs of time associated with a particular aspect of recreation than the wage rate which measures the trade-off between work and leisure, generally. The results seem to indicate a better overall fit for the models with the elicited value of individual consumer's time than for the models with the more traditional hourly earnings (wage rates). The importance of the correct measurement of the opportunity cost time is illustrated by showing that estimated consumer surpluses based on two different value of time measurements differ significantly.

Abstract: Over the last few decades increasing concern has been voiced regarding the detrimental effect which many modern farming practices are thought to have on the countryside. Environmentally Sensitive Areas (ESA) Scheme is one response to this concern and seeks to support those agricultural practices which have contributed to the distinctive agricultural landscape, which for many people typifies the British countryside. This paper uses the contingent valuation method to estimate the benefits which the general public receives from the ESA scheme in the South Downs and sets them against the estimated net exchequer costs of the scheme. The study illustrates the importance of including passive use values in benefit-cost analysis, if environmental protection is to be provided in certain marginal areas.

Abstract: Much interest now focuses on the use of the Contingent Valuation Method (CVM) to assess non-use value of environmental goods. The paper reviews recent literature and highlights particular problems of information provision and respondent knowledge, comprehension and cognition. These must be dealt with by economists in designing CVM surveys for eliciting non-use values. Cognitive Questionnaire Design Methods are presented which invoke concepts from psychology and tools from cognitive survey design (focus groups and verbal reports) to reduce a complex environmental good into a meaningful commodity that can be valued by respondents in a contingent market. This process is illustrated with examples from the authors' own research valuing alternative afforestation programmes.

Abstract: Benefit transfer is defined and the reasons for its growing popularity are considered. The various approaches to benefit transfer are analyzed and existing methods of assessing its accuracy are supplemented with reference to the literature on the reliability of contingent valuation estimates.

Abstract: Fourteen papers, one previously published, resulting from a conference held at St. Aidan's College, Durham, U.K., in March 1994, examine economic and ecological perspectives on environmental values. Papers focus on the assumptions of cost benefit analysis, a philosopher's view of the assumptions of cost-benefit analysis; the role of environmental valuation in cost-benefit analysis; ecological and economic values; economic valuation and ecological values; contingent valuation and economics; contingent valuation in a policy context; alternative valuation techniques; the pros and cons of alternative valuation methods; uncertain values in environmental policy analysis; transferability of benefit estimates; a willingness-to-pay based value of underground safety; establishing priorities for valuation research in government; and value for money in environmental valuation. Willis is at the University of Newcastle upon Tyne. Corkindale is with the Department of the Environment, London.

Abstract: Subjects valued environmental goods using a response mode expressing willingness to pay as a multiplier or divider of a purported 'budgetary unit.' Hypothetical willingness to pay was found to be highly dependent on the size of the budgetary unit. Rejecting two other interpretations, the authors believe the results suggest that people's contingent values are only vaguely represented in monetary terms; any dollar amount within a broad range is considered an acceptable expression of value. Although the procedures differ from those of a typical contingent valuation survey, the results demonstrate some of the difficulties in discovering precise monetary values for environmental resources. Coauthors are Sarah Lichtenstein, Thomas C. Brown, George L. Peterson, and Paul Slovic.

Abstract: Eight papers discuss the contingent valuation method and experimental methods that can be used to value environmental assets, and show how altruistic concerns can be handled in evaluations of environmental change. Papers focus on current issues in the design, administration, and analysis of contingent valuation surveys; the interpretation of responses in contingent valuation surveys; preference uncertainty, optimal designs, and spikes; valuing changes in health; an experimental economics perspective on whether real environmental benefits can be estimated; current issues in resource accounting; environment-economy interactions in a computable general equilibrium model of Sweden; and intertemporal equilibrium modeling of energy and environmental policies. Contributors are mainly economists. Johansson and Kristrom are at the Stockholm School of Economics. Maler is at the Beijer Institute of Ecological Economics.

Abstract: Two Monte Carlo willingness-to-pay (WTP) models were constructed to compare three approaches that generate bids for the referendum CVM. In the first model, WTP was normally distributed, and in the second, was log normally distributed. The bid approaches were those of K. L. Boyle, M. P. Welsh, and R. C. Bishop, (1993), and an ad hoc third approach. Some properties of these approaches were discussed. WTP estimates from the three approaches were compared to the true value. Results indicated that Boyle et al.'s approach was preferred, especially when variation in WTP was low. Estimates from the three approaches became more comparable as WTP variability and sample size increased.

Abstract: The empirical literature on discrete response contingent valuation has found that seemingly innocuous changes in the statistical models estimated result in significantly different point estimates of willingness to pay. This paper hypothesizes and tests several potential explanations for these results. First it investigates and compares the biases inherent in single-bounded and double-bounded maximum likelihood estimation procedures and examines how they react to various bid designs and sample sizes. Then it examines the presence and identification of "outliers" in binary choice data and how these outliers influence estimation. Finally, it presents an alternative approach to addressing the issue of outliers which explicitly acknowledges the possibility of upwardly biased response probabilities.

Abstract: The term "embedding" is ill-defined and has been applied to distinct phenomena, some predicted by economic theory and others not. This paper lays out a theoretical framework for looking at these phenomena and provides a set of well-defined terms. Included is a discussion of survey design problems which may induce spurious evidence in favor of the hypothesis that respondents are insensitive to the scope of the good being valued. An empirical example of the component sensitivity is provided. This test rejects the hypothesis that respondents are insensitive to the scope of the good being valued.

Abstract: In 1991, the U.K. Department of Transport commissioned a nationally representative sample survey with a view to estimating willingness-to-pay based monetary values for the prevention of nonfatal road injuries. The study design was somewhat unusual in that it involved not one but two main value elicitation procedures--contingent valuation and standard gambles. Although both procedures are rooted in the same conventional theoretical foundations, they produced substantially and systematically different empirical estimates of the values being sought. This paper reports the findings of the study and examines possible reasons for the disparities referred to.

Abstract: Considers the evaluation issues faced by health economists planning to use the willingness-to-pay approach in their work. Designed as a graduate-level textbook and as a reference text for those undertaking evaluations of projects involving health risks. Presents basic tools and definitions. Introduces money measures of the value of changes in health care; examines their properties in a certain world; and generalizes the money measures for the case involving risk. Addresses the problem of preference revelation, discussing the main methods used, or suggested, for measuring willingness-to-pay for health services--survey techniques, the human capital approach, and indirect methods using market data. Surveys empirical studies using different methods to assess the value of health changes and the value of a statistical life. Reviews studies using the contingent valuation method to value health care and discusses the steps in a contingent valuation method study of willingness-to-pay for a medical treatment. Examines the aggregation problem of summing willingness-to-pay across individuals in order to decide whether a project is socially profitable. Addresses further evaluation issues involving the value of information, the health production function approach, and the value of changes in life expectancy. Briefly discusses the outcome measures typically used by health economists--quality-adjusted life-years and healthy-years equivalents--and considers their usefulness in decision-making.

Abstract: Test-retest, the standard method used to investigate reliability of contingent valuation, is limited because when the time period between the two surveys is relatively short the study may exhibit testing recall. Conversely, when the time between the two surveys increases there is an increased chance the true value will change. The test-retest design cannot isolate these effects. Independent pretest and posttest control groups allow testing of reliability even when recall occurs or the true value changes. Using this design, the authors found ex post estimates of Hicksian surplus are reliable regardless of whether respondents have direct experience with an activity. Coauthors are Kevin J. Boyle, Daniel W. McCollum, and Stephen D. Reiling.

Abstract: The consistency of an individual's willingness to pay (WTP) responses for increases in the quantity of an environmental public good (whale populations) is tested along three lines. First, we test whether WTP for 50 percent and 100 percent increases in whale populations are statistically different from zero. Second, we ask whether the incremental WTP from a 50 percent increase to a 100 percent increase is statistically significant. Finally, we test whether there is diminishing marginal valuation of the second 50 percent increment in gray whale populations. The paired t-tests on open-ended WTP responses supported all three sets of hypotheses. Both visitors and households provided WTP responses that were statistically different from zero and increased (but in a diminishing fashion) for the second increment in WTP. In this survey both visitors and households provided estimates of total economic value (including non-use or existence values) for large changes in wildlife/fishery resources that were consistent with consumer theory.

Abstract: Two different willingness-to-pay responses are compared--willingness to pay to avoid loss of air quality and willingness to pay to obtain gains in air quality. Contingent valuation data are used to estimate bid functions for these two types of responses for visibility and health. Comparison of the estimated models indicates that, in addition to magnitude differences, gains and losses for visibility and health are affected differently by health status, risk perceptions, and other risk-related variables. Results for the loss measure are more reliable in terms of variability of response and econometric modeling.

Abstract: A discrete/continuous model of participation in environmental programs is developed. A conditional expectation correction model is used to overcome the selectivity bias inherent in single-equation approaches used to describe participation. The two-equation model proposed explicitly recognizes the linkage between the discrete decision (participate or not) and the continuous decision (level of participation for enrollees). This approach yields a measure of willingness to accept payment for program participation that is not obtainable when level of participation alone is considered. A model using contingent valuation survey data assessing participation in the filter strip program in Michigan serves as an example.

Abstract: Pricing of road transport at social marginal cost should make users aware of the cost of damaging the environment (external costs). such an approach, however, requires a monetary estimate of this damage which can be difficult to derive. On the basis of a study carried out at Neuchatel (Switzerland), the contingent valuation method was used to estimate the damage caused to buildings of historical and cultural value by traffic-caused air pollution. In a survey, individuals were asked to contribute to a fund set up to finance the maintenance of pre-selected historic buildings. A valuation function to predict willingness-to-pay responses is estimated.

Abstract: This article compares willingness-to-pay (WTP) estimates from an actual survey using a checklist question regarding WTP for groundwater quality improvements to WTP estimates that would have been obtained had a single-bounded referendum (SBR) or a double-bounded referendum (DBR) question been asked. Results indicate differences among estimates from the three types of question formats. There was a loss of statistical efficiency of parameter and WTP estimates when moving from the checklist and DBR formats to the SBR format. WTP estimates from the SBR question were more sensitive to sample size and model specification than the others.

Abstract: In this paper, the empirical implications of altruism for cost-benefit analysis of projects involving health changes are investigated. It is shown that a willingness-to-pay question allowing the respondent to state her total willingness-to-pay (irrespective of what reasons she may have for paying), subject to everybody else paying so as to stay at their initial levels of utility, produces, as a special case, the project evaluation rules derived by M. W. Jones-Lee and others. The implications of alternative formulations of the valuation question in a contingent valuation study are also explored.

Abstract: This paper is based on the work of the Resource Assessment Commission, which was required by the Prime Minister to inquire into uses of Australia's forest and timber resources. It begins with a brief discussion of concepts of sustainable resource management and the guiding principles of the Commission. A description is given of the study area, its environment and forest resources. Sources of conflict in resource management are identified. The paper then explains how benefit-cost analysis techniques were applied to assess resource management options. The analysis undertaken was based on the Krutilla-Fisher model and contingent valuation of the forest resource.

Abstract: Contingent valuation surveys have become an important tool in placing monetary values on non-market goods and amenities. Many policy issues involve evaluation of several alternatives such as different environmental quality levels, different levels of risk, etc. Contingent valuation then involves asking the respondent a sequence of nested questions. Asking and analyzing a nested sequence of questions is an efficient approach to data gathering and preference revelation. The resultant sequentially censored data set cannot be efficiently analyzed with the standard regression models like the Tobit or nested logit models. The nested Tobit model is proposed as an efficient and consistent method of estimating regressions using sequentially censored data. An empirical application suggests greater efficiency in comparison to the Heckman two-stage procedure.

Abstract: Reliability is an important dimension of urban water supply. While out-of-pocket losses appear small, water customers place high values on reliability. This demand for reliability can be measured by contingent valuation methods, while the costs of providing various levels of reliability can be estimated through hydrologic simulation. A framework for optimizing reliability is presented and the contingent valuation survey is described. Results from three Colorado towns are presented. 1994 Academic Press, Inc.

Abstract: Month Carlo experiments were conducted to compare regression parameter and willingness-to-pay estimates using the contingent valuation method's referendum and payment-card questioning formats. Strategic bias was controlled by making responses consistent with the true willingness to pay. Two willingness-to-pay models were constructed: linear and log-linear in explanatory variables. Based on the efficiency and mean-squared-error criteria, the payment-card model was superior to the referendum model. However, neither model outdid the other when the unbiasedness criterion was used. The performance of either model depends on the choice and distribution of the bids and payment-card categories.

Abstract: Contingent valuation of people's willingness to pay has rapidly become the method of choice to value all manner of environmental damages. The correct measure is, however, the sum people require to compensate them for such losses, an amount which will normally be far larger than their willingness to pay. And on present evidence, responses to contingent valuation questions are not likely to represent any measure of economic values. The results of these valuation practices will, therefore, bias environmental policies and distort incentives.

Abstract: The Australian Resource Assessment Commission conducted a large contingent valuation study to estimate the economic benefits of preserving Kakadu Conservation Zone (KCZ) by incorporating it into Kakadu National Park rather than allowing mining in it. Different subsamples were presented, with scenarios representing the environmentalist (major) and mining industry (minor) views of the impacts. Willingness to pay to prevent the major impact scenario was twice as high for the minor impact scenario. For both scenarios, benefit-cost analysis indicates that preservation should be chosen over mining. A valuation function to predict the willingness to pay responses is estimated.

Abstract: Willingness-to-pay responses from dichotomous choice contingent valuation studies are often modeled using logistic regression, from which estimates of mean or median willingness-to-pay are calculated. However, a great many factors influence an individual's willingness-to-pay, some of which may be unobserved. Hence, the regression model may have inadequate explanatory power, and parameter estimates may be biased and their significance overestimated. The effects of this overdispersion are examined within a generalized linear mixed modeling framework and an example given using published data from J. Cooper and J. Loomis (1992).

Abstract: The question at issue is this: can we, as social scientists, measure in monetary terms the value that people place on the natural environment, or other public goods, when there is no other corroborating market behavior that would allow measurement by revealed preference methods? One of the great errors of analysis is the fallacy of misplaced precision meaning the wrong thing with exquisite precision, rather than the right thing with lesser accuracy. It certainly is true that no number can sometimes be better than some number. Is it true that no number must always be better than some number? Is it, true, as Diamond and Hausman (DH) assert, that CV surveys never measure people's preferences and are never a suitable source of information on values in either benefit cost analysis or damage assessment? DH take the position that ordinary people are hopelessly ill informed, liable to exaggerate the damages from oil spills, and should not be trusted to make judgements concerning monetary values. Empirical demand analysis traditionally dealt with a representative consumer and broadly defined commodities. This was driven by the type of data available, which were generally highly aggregated over both individuals and commodities. Yet our observation of economic life shows that this is too simplistic. Contingent Valuation provides one way in which the more realistic theory of individual behavior can be given practical expression. It provides one way of tracing out the demand curve for a commodity that cannot be revealed through market data, but nevertheless exists and should count in an economic analysis.

Abstract: Contingent valuation is now used around the world to value many types of public goods, including transportation, sanitation, health, and education, as well as the environment. The author describes how researchers go about making such surveys reliable, mentioning recent innovations in sampling, questionnaire design, and data analysis, including formulating the valuation as a closed-ended question about voting in a referendum to raise taxes for a particular purpose. He addresses various objections that contingent valuation results are incompatible with economic theory. Even without a market, there still exists a latent demand curve for nonmarket goods; contingent valuation represents a way to tease this out.

Abstract: A test is performed of the recommendation by the National Oceanic and Atmospheric Administration's panel on contingent valuation that respondents be reminded about other substitute resources and their budget constraint prior to answering the willingness-to-pay question. The context for this experiment is a contingent valuation method study of the benefits from reducing fire hazards to old-growth forests in Oregon. The test was performed using two mail surveys that were otherwise identical except for the inclusion of reminder statements prior to the willingness-to-pay question. The dichotomous choice logit equations were not statistically different between versions and the mean willingness-to-pay results were identical.

Abstract: From a study of eighty-two households in urban areas of Tunis, Tunisia, it appears that the most important obstacle in connecting to the piped water system is the utility-required cash down payment. Results from the study's contingent valuation questions showed that the respondents could easily afford the anticipated monthly charges for piped water and sewer service, and their bids for this service were very close to those for households already receiving them. A further conclusion is that, as the reliability of the piped water system increases, then the importance of a monthly tariff in the connection decision decreases.

Abstract: In this paper, the authors address specification bias in contingent valuation by determining the expected effects of omitting relative price variables on willingness-to-pay and including measures of own-price and cross-price in an empirical contingent valuation function. They find that omission of an own-price variable downwardly biases the income coefficient. Omission of a cross-price variable results in an upwardly biased income coefficient. Omission of the cross-price variable also upwardly biases the own-price coefficient. These results show that omission of relative prices from willingness-to-pay equations may lead to biased econometric results.

Abstract: This paper compares the performance of four approaches to calculating confidence intervals around dichotomous choice contingent valuation method benefit estimates. The performance of the approaches is compared using Monte Carlo simulation techniques for the two most common specifications for the welfare estimate. The results indicate that all four methods tended to perform well on average but the methods differed in the frequency with which they performed well. The results indicate the best choice depends on the sample size, on the distribution of the welfare estimate, and on the choice of functional form for the welfare estimate.

Abstract: This paper investigates what a respondent is thinking when answering a willingness-to-pay question in a contingent valuation, using a "think aloud" technique from psychology called verbal protocol analysis. The willingness-to-pay responses we observed seem to be constructed from a variety of considerations, including an obligation to pay a fair share of the cost of the solution and signaling concern for a larger set of environmental issues. The finding that respondents seem to construct their values at the time they are asked, rather than reporting a more well-defined value, is seen as consistent with over two decades of research on the psychology of decision making. Potential uses of verbal protocols in contingent valuation studies are also discussed. 1994 Academic Press, Inc.

Abstract: The District Court of Appeals for Washington D.C. has upheld a Department of the Interior ruling that accepts as a best available procedure for reliably calculating natural resources damages the Contingent Valuations Method (CVM). This paper addresses a number of questions that are relevant for assessing the reliability, or accuracy, of CVM values. Empirical evidence concerning the extent to which CVM subjects behave strategically, i.e., misrepresent their true values, is found to be mixed. While results from received empirical studies do not support unequivocal conclusions one way or another, their results suggest that CVM values may overestimate real economic commitments by a considerable margin. It is then argued that the courts cannot accept CVM values out of hand; they must carefully assess the demonstrated potential for such overestimates of value in their deliberations concerning the use of results from CVM surveys in CERCLA litigation.

Abstract: This Article evaluates existing literature on the performance of contingent valuation. It considers: the performance of the methods in natural resource damage assessments; the implications of the Ohio court decision for the use of CVM; the relationship between laboratory experiments on economic valuation and CVM; and the implications of the NOAA Panel's recommendations.

Abstract: The Australian Resource Assessment Commission conducted a large contingent valuation study to estimate the economic benefits of preserving the Kakadu Conservation Zone by incorporating it into Kakadu National Park rather than allowing mining in it. Different subsamples were presented, with scenarios representing the environmentalist (major) and mining industry (minor) views of the impacts. Willingness to pay to prevent the major impact scenario was twice as high as for the minor impact scenario. For both scenarios, benefit-cost analysis indicates that preservation should be chosen over mining. A valuation function to predict the willingness to pay responses is estimated.

Abstract: Existence values are playing an increasingly important role in wildlife preservation decisions but little is known about how these values behave over time and value estimates may often be misinterpreted. Panel study results suggest that, although existence values may be relatively stable, many individuals respond to contingent valuation by bidding their fair share or for the satisfaction from contributing to a good cause, such as environmental quality. Although payment of fair share may represent a lower bound estimate of existence value, payment for a good cause may or may not be closely related to the value of the resource itself.

Abstract: This paper describes the results of a test of the contributions model for respondents' answers to contingent valuation (CV) questions. The test uses a random sample with independent assignment of respondents to one of two different treatments that describe plans to either improve the landscape on interstate highways or support recycling of used automobile and truck tires. A discrete choice framework with the same fee structure was used for each plan. Four tests based on respondents' choices all suggest significantly different responses to the two plans. Moreover, all relevant economic variables were found to be statistically significant determinants of stated choices. Thus, for these two plans, the results indicate CV is able to distinguish respondents' choices to isolate significantly different implied economic values for each plan.

Abstract: A traditional contingent valuation approach and the "trip response method" were examined as potential techniques for measuring public area recreation revenue-capture potential. Empirical results suggest that both methods are useful for assessing revenue-capture potential. Additional research on alternative methods for assessing recreation revenue-capture potential is encouraged.

Abstract: Originally prepared as the basis for a shorter talk on emerging issues in resource and environmental economics for the 35th anniversary of Resources for the Future, this paper has been updated and expanded to include current literature including discussion of the controversy over the contingent valuation for estimating the economic value of environmental resources.

Abstract: Contingent Valuation has been promoted as a catch-all approach to environmental valuation. While there have been numerous attempts in recent years to place monetary values on environmental amenities, studies have often reported a high frequency of protest, zero or inordinately large dollar-value responses. This paper reports on the results of a survey designed to obtain information on how people actually interpret questions of paying to avoid changes in their views of Rangitoto Island. Evidence suggests that the meaning respondents attach to the actual dollar values they offer or bid are inconsistent with the conventional logic that underlies Contingent Valuation. Instead, respondents might be seen to be expressing views about how things ought to be in society, and that it is simply not right to develop Rangitoto Island.

Abstract: There is increasing recognition that statistical estimates of the average willingness-to-pay for environmental programs may be biased by the failure of respondents, or of the statistical methodology, to consider the effect of substitute programs and goods. The authors design and implement a contingent valuation method survey that estimates the impact of substitution effects on the willingness-to-pay for an environmental program. They find a statistically significant impact and suggest that the set of potential substitutes is very large.

Abstract: The justification of a subsidy relies upon the museum's social value which consists in its use value, its non-use value, and in its externalities. The estimation of these values is based or the consumer sovereignty expressed by the net cash revenues of the museum, the consumer surplus, positive externalities and the willingness to pay taxes. The paper measures these sources of value by the travel cost and contingent valuation methods and by special computations of externalities. The particular museum, used in this paper as a case study, seems to pass the test, i.e, its subsidies do not exceed its social value.

Abstract: Without market outcomes for comparison, internal consistency tests, particularly adding-up tests, are needed for credibility. When tested, contingent valuation has failed. Proponents find surveys tested poorly done. To the authors' knowledge, no survey has passed these tests. The 'embedding effect' is the similarity of willingness-to-pay responses that theory suggests (and sometimes requires) be different. This problem has long been recognized but not solved. The authors conclude that current methods are not suitable for damage assessment or benefit-cost analysis. They believe the problems come from an absence of preferences, not a flaw in survey methodology, making improvement unlikely.

Abstract: The significance of the court's decision in Ohio v. Interior Dept. is easy to overstate, since the court held only that the Interior Department had not acted capriciously in including contingent valuation methodology (CVM) among several alternatives available for the conduct of natural resources damage assessments (NRDAs). In practice, many procedural and institutional safeguards will protect us from erroneous NRDAs. In any event, the entire NRDA process is overly complicated, virtually impossible to administer and of uncertain utility. Consideration should be given to abandoning it altogether, or at least to replacing it with a surcharge on response and remediation costs.

Abstract: The contingent valuation method for estimating the existence value of natural resources is examined for psychophysical robustness, statistical reliability, and economic sensibility. Extensions of standard models for willingness-to-pay, and suitable econometric techniques for analyzing these models, are developed. The analsis is applied to a series of experiments on the value of preserving wilderness areas in the western United States. The results call into question the reliability of the contingent valuation method for estimating existence values.

Abstract: A literature search provides 83 studies from which 616 comparisons of contingent valuation (DV) to revealed preference (RP) estimates are made. Summary statistics of the CV/RP ratios are provided for the complete dataset, a 5% trimmed dataset, and a weighted dataset that gives equal weight to each study rather than each CV/RP ratio is 0.89 with a 95% confidence interval (0.81 0.96) and a median of 0.75. For the trimmed and weighted datasets, these summary statistics are (0.77,(0.74-0.81);0.75) and (0.92;[0.81-1.03];0.94), respectively. The Spearman rank coalition coefficients between the CV and RP estimates for the three datasets are 0.78, 0.88, and 0.92, respectively, with the Pearson correlations a bit larger. Non-parametric density estimates are provided, as well as the results of regressions of the observed CV/RP ratios on the basic RP technique used and the broad class of goods valued.

Abstract: This study assesses the amount individuals voluntarily contribute when visiting a Cathedral where no charge is made for entry. Contingent valuation methods are used to estimate the maximum individuals would be willing to pay if an entry charge was imposed, the changes in visit rates which would ensue at different price levels and the maximum revenue the Cathedral could be expected to collect by charging an entry fee.

Abstract: The contingent valuation method, wherein sample surveys are used to elicit individuals' willingness to pay for certain types of policies, is playing an important role in government decision-making. The most prominent applications are in the valuation of damages to natural resources from oil spills. But the contingent valuation method will be more important still if it is used to expand the range of impacts included in applied benefit-cost analyses. This paper explains the origins of the contingent valuation method and the route through which it came to be the center of both an academic and a legal debate.

Abstract: Resampling or simulation techniques are now frequently used in applied economic analyses. However, significance tests for differences between empirical distributions have either invoked normality assumptions or have used nonoverlapping confidence interval criteria. The authors demonstrate that such methods generally will not be appropriate and they present an empirical test, based on the method of convolutions, for assessing the statistical significance between approximate empirical distributions created by resampling techniques. The proposed convolutions approach is illustrated in a case study involving empirical distributions from dichotomous choice contingent valuation data.

Abstract: Agricultural land often yields significant amenity benefits that may not be reflected in the prevailing allocation of land between agricultural and nonagricultural uses. This paper introduces a simple supply-demand model that incorporates these benefits in the determination of optimal land allocation. Three-stage least squares and county-level data are used to estimate the supply and demand parameters. These estimates confirm the plausibility of the conceptual model. Results from the contingent valuation literature illustrate the extent of underallocation of land to agriculture due to existing institutional conditions. Some corrective measures and their distributional effects are discussed and quantified.

Abstract: A theoretical framework for understanding substitution and complementary effects between environmental programs is derived. A multiple program valuation function is specified and its structure analyzed. A maximum likelihood estimator is derived to estimate the parameters of the valuation function using referendum-type, contingent valuation data. The empirical analysis shows that the conventional benefit aggregation approach of independent valuation and summation overstates the benefits of two- and three-program policies by an average of 24 percent and 54 percent respectively.

Abstract: This paper is about a consistency check suggested by the recent NOAA Blue Ribbon Panel as one basis for evaluating contingent valuation (CV) surveys. This check relates to the responsiveness of valuation estimates to changes in scope or size of the environmental commodity. This paper challenges the claims that the CV method is incapable of demonstrating sensitivity to scope and is therefore too unreliable to obtain useful information for natural resource damage assessments. These claims are questioned on the basis of evidence from the larger CV literature and a re-analysis of the studies conducted by the critics. Findings indicating that CV is sensitive to issues of scope.

Abstract: Most of the work on valuing the recreational benefits of public forests has concentrated on estimating consumers' surplus per visit. We use two contingent valuation approaches to try to place a value on the physical characteristics of British forests, and hence to explain the variation in consumers' surplus across different forest types. Our two approaches are, first, to obtain incremental willingness-to-pay to access forests with different levels of a number of characteristics by showing visitors pairs of photographs; and second, to use characteristic levels as explanatory variables in a bid curve. We find that forest characteristics are in general poor predictors of willingness to pay.

Abstract: Eleven papers, plus comments, resulting from a conference organized by Cambridge Economics, Inc., and held in Washington, D.C., in April 1992, offer a critical assessment of the survey method known as contingent valuation. Papers focus on contingent valuation-based estimates of nonuse value; the properties of contingent valuation-based measures of nonuse value; user damages in the case of an oil spill; and economic and legal problems using contingent valuation-based measures of nonuse value.

Abstract: The use of contingent valuation (CV) methods for estimating the economic value of environmental improvements and damages has increased significantly. However, doubts exist regarding the validity of the usual willingness to pay CV methods. In this article, we examine the CV approach in light of recent findings from behavioral decision research regarding the constructive nature of human preferences. We argue that a principal source of problems with conventional CV methods is that they impose unrealistic cognitive demands upon respondents. We propose a new CV approach, based on the value-structuring capabilities of multiattribute utility theory and decision analysis, and discuss its advantages and disadvantages.

Abstract: The prevailing consensus that existence value can only be measured by contingent valuation is questioned. The concern is over 'pure' existence value, a change in welfare when a public good changes that leaves no trace in behavior. Pure existence value may well be more a consequence of the simple models used to characterize nonuse value than a reflection of how people value public goods. If so, nonuse value can, in principle, be obtained from observing behavior as well as from asking questions, which offers a prospect for cross-checks and corroboration using both approaches. Some possibilities for empirical measurement are suggested.

Abstract: Recent developments related to the use of contingent valuation (CV) are discussed, especially the debate in the United States over the use of CV in natural resource damage assessment. It is argued that two misconceptions underlie much of the current debate over contingent valuation. After an analysis of these misconceptions and their implications the paper considers some issues in the design, administration and analysiso of reliable contingent valuation surveys.

Abstract: Natural resource damage liability (NRDL) has changed the role of economics for decisions involving natural and environmental economics. The statutes defining NRDL recognize these resources as natural assets. Monetary measures of their value are central in determining the liabilities facing parties held responsible for injuring them. This paper describes the issues raised by NRDL for: translating injuries into economic commodities, the use of contingent valuation, economic issues raised by restoration planning, and conventional evaluations of the incentive properties of liability rules.

Abstract: Binary choice models are frequently used in the analysis of referendum contingent valuation data. Because omitted regressors, heteroskedasticity, and distribution asymmetry in these models result in inconsistent parameter estimates, it is important that misspecification tests be undertaken. In this article, conditional moment tests for these problems are presented and applied to data from two referendum contingent valuation studies. Results show that some models have misspecification problems and that these problems affect the estimation of welfare measures.

Abstract: This paper is part of a compendium discussing what we actually know about the performance of the contingent valuation method for measuring the values people place on nonmarketed resources. After describing the strengths and weaknesses of alternative tests for the consistency between CVM responses and the predictions of economic theory, the evidence from laboratory and field experiments is reviewed. There is also a summary of the findings from two meta analyses of CVM estimates. The paper concludes with a review of the compatibility of the recommendations of the NOAA Panel on Contingent Valuation in relation to the available evidence on the method's performance.

Abstract: This paper examines procedures based on expressions of willingness to pay to determine the provision of public goods such as environmental amenities. In addition to classical efficiency and majority political support criteria, some issues related to the practical implementability of such mechanisms are considered, including the incentives agents have to report truthfully and the possibility of sampling a subset of the population of agents. In particular the paper suggests a decision procedure that is an alternative to those based on estimating either the mean or median willingness to pay, as proposed in conventional Contingent Valuation studies. Based on an average of conditional median values of willingness to pay, it is shown that this mechanism has many desirable properties. While maintaining incentives for agents to report truthfully their valuation for a public good, the mechanism limits the effects of outliers and broadens the political acceptability for adopted policies. In addition, the mechanism has better statistical properties than conventional Contingent Valuation procedures when only a sample of individuals is queried for their willingness to pay.

Abstract: The author compares the informational efficiencies of contingent rating, contingent ranking, and two contingent paired-comparison methods as alternatives to the referendum contingent valuation method. The contingent rating method is hypothesized to be most efficient because ratings convey information on preference intensities and can uniquely represent respondent indifference or ambivalence. Survey data on hunters' ratings of alternative hypothetical hunting trips are used to estimate four alternative indirect utility models from which marginal willingness-to-pay measures for individual trip attributes are derived. Model comparison, willingness-to-pay estimates, and their confidence intervals confirm the relative efficiency of the contingent rating approach.

Abstract: The double-bounded dichotomous choice model is becoming popular in the contingent valuation literature. This technique considers the responses to two bids, the second one being determined according to the response to the first, and it has been shown to be statistically more efficient than the standard, single-bounded model. In this paper, the issue of the optimal choice of bid values for the double-bounded logit model is addressed. These designs are derived to satisfy the D-optimal, C-optimal, and fiducial method criteria.

Abstract: This study has proved that without taking into account the substitution and complementarity relationships among policy components, the economic benefits of a complex policy package can be miscalculated. The conceptual framework has been tested by undertaking an empirical study, using a contingent valuation survey, and is in regards to the regulatory and management policy agenda in central Ohio's Big Darby Creek (one of the most biologically diverse creeks found in the mainland United States). Our results provide strong evidence that substitution effects exist among the components of the complex policy under evaluation. On average, the independent valuation and summation of benefits for all environmental attributes results in calculated total benefits about twice as great as the value estimated with sequential valid total valuation procedure.

Abstract: Sequential experimentation is recommended as an alternative to the usual one- or two-shot approach to contingent valuation surveys. The sequential C-optimal approach is introduced and developed, and it is compared to the established parametric and non-parametric sequential procedures. It is shown to be an improvement in terms of statistical precision in a finite sample, over the standard sequential procedures.

Abstract: A test is developed to find whether open-ended and closed-ended contingent valuation mechanisms lead to significantly different results. The test is based on joint estimation of willingness-to-pay responses to open- and closed-ended questions asked of the same sample of individuals. In a public good example, individuals do respond differently depending on question format. Possible explanations include different incentives for strategic behavior and respondents' lack of familiarity with the open-ended question type. No differences in willingness-to-pay were found in a private good example.

Abstract: We test for an embedding effect on a geographically distributed public good and the extent to which improvements in the CVM design reduce the occurrence and magnitude of embedding effects in valuation. Using both open-ended and dichotomous choice CVM for protection of forests in all of southeastern Australia and two smaller portions of that area, we find the occurrence of embedding effects in only one of the two levels. When embedding effects were present, there was a reduction in value much smaller than that found by Kahneman and Knetsch.

Abstract: Two versions of a contingent valuation questionnaire that differ only in the device used to communicate risks from hazardous waste were sent to California households. Both risk communication devices provide willingness-to-pay functions that vary in a statistically significant fashion with the absolute risk reduction. However, these empirical results also demonstrate that different risk communication devices produce statistically different logit equations and, hence, estimates of willingness to pay for reductions in exposure to hazardous waste. The pros and cons of using each risk communication device in contingent valuation method surveys is also discussed.

Abstract: We use a combination phone/mail survey to test for possible sample biases in contingent valuation. We find no sample selection bias but do find non-response bias. We show how failure to correct for non-response bias distorts aggregate benefit estimates.

Abstract: Benefit-cost analysis of coastal and marine wildlife management programs requires that economic benefits be monetized for comparison with the costs of preservation. Without explicit measurement and consideration of nonuse values, benefits may be underestimated and resources devoted to wildlife programs may be underallocated. Using data from a contingent valuation survey of nongame wildlife programs in coastal North Carolina, this paper provides additional evidence that total economic values under uncertainty for wildlife are theoretically valid. Specification error is found for valuation models which do not include measures of uncertainty. Specification error can lead to errors in benefit estimation.

Abstract: Fourteen papers examine the economic valuation of the natural environment, with particular emphasis on Europe. Papers focus on a review of European valuation studies and their use; the validity of contingent valuation surveys; and decisionmakers' use of valuation studies. Contributors are mainly economists.

Abstract: This chapter gives an overview of the techniques used to value the non-market benefits of water-related public goods and of the major empirical studies in this area. Travel cost, hedonic pricing, and contingent valuation are described; special emphasis is placed on the problems and limitations of implementing these methods to value changes in the quality and quantity of water-related amenities. Major empirical efforts to value national and regional water quality improvements, water-based recreation, ecosystem preservation, instream flows, groundwater protection and water supply reliability are discussed.