Contingent Valuation Method

Contingent Valuation Method: Concept Architecture

Orientation and learning roadmap

This page follows a stated-preference survey from its hypothetical health scenario to a monetary valuation. It explains how the payment question is designed, works through a simple willingness-to-pay aggregation, and examines validity and equity issues before the result is used in policy analysis.

Describe the change being valued

Specify the starting situation, the proposed health service or outcome change, who receives it, how long it lasts, and what remains uncertain. Respondents need a credible alternative so that their answer refers to the same improvement rather than to different imagined versions of it.

The payment vehicle also matters: an increase in tax, insurance premium, or personal charge can elicit different responses because each implies a different distribution of costs and benefits. State whether the survey seeks willingness to pay for a gain or willingness to accept compensation for a loss; these need not be equal. A methodological assessment of contingent valuation in healthcare identifies scenario, payment vehicle, risk presentation, and timing as central design choices.

Ask for a monetary trade-off

An open-ended question can request the largest amount a person would pay for the specified change. A referendum-style question instead asks whether the person would support the change at a stated cost, with different respondents offered different bids to learn about the distribution of willingness to pay.

Those formats require different analyses: a yes response at a bid establishes that the person's stated value meets at least that bid under the scenario, not their exact maximum. Survey instructions should make the budget constraint and available alternatives clear, and responses such as protest answers should be handled transparently. The NOAA Panel report on contingent valuation discusses careful scenario description, referendum questions, and survey checks.

Aggregate values only for the intended population

For an open-ended survey with defensible individual values, a sample mean is one possible summary of stated willingness to pay. To estimate a population total, that mean must represent the population that would actually receive or fund the change, and the proposed payments must not value the same benefit twice.

If $w_i$ is respondent $i$'s stated maximum payment and the sample represents $N$ eligible people, a simple illustration is:

$$\overline w=\frac{1}{n}\sum_{i=1}^{n}w_i, \qquad V=N\overline w.$$

Survey weights, nonresponse, uncertainty, and household versus individual payment units may change this calculation. For referendum data, estimate the share willing to pay at each bid and infer a willingness-to-pay distribution under explicit model and range assumptions instead of averaging the offered bids as if they were personal maxima.

Worked willingness-to-pay example

Suppose four illustrative respondents state maximum one-time payments of 20, 40, 60, and 80 currency units for a clearly described service. Their simple mean is 50 units, and applying it to 1,000 equally eligible people gives an illustrative aggregate value of 50,000 units.

If provision would use resources costing 40,000 units on the same time and valuation basis, the arithmetic difference is 10,000 units. Four respondents cannot establish a representative population mean, and the calculation does not demonstrate that the service should be funded. Real analysis must examine sampling, payment incidence, uncertainty, public budget constraints, and any benefits or costs not captured by the stated scenario.

Test whether responses reflect the intended value

Hypothetical answers may differ from payments people would actually make. Starting bids, ranges displayed on payment cards, strategic responses, unfamiliar risk descriptions, and failure to distinguish small from large improvements can affect stated values.

Pilot the survey and test understanding, plausibility of the payment vehicle, sensitivity to the scope of benefit, and sensitivity to income or ability to pay. The healthcare contingent-valuation review documents methodological and conceptual concerns, including bid-range effects. A high response rate or precise statistical estimate does not by itself resolve bias in what was asked.

Interpret values for health economic decisions

Contingent valuation can express a specified health-related benefit in money for a cost-benefit analysis when the survey question and target population are defensible. It is not interchangeable with quality-adjusted life-years or a health-system cost-effectiveness threshold, and willingness to pay can reflect ability to pay as well as preferences.

Report whose preferences were elicited, whose budget was invoked, the time horizon, benefit definition, survey method, monetary year, and aggregation population. Examine distributional implications rather than allowing high-income respondents to dominate an equity judgement solely because they can state larger payments. Use the estimates alongside clinical evidence and resource costs, not as a substitute for them.