A Review of Methodological Issues in the Conduct of Willingness-to-Pay Studies in Health Care I: Construction and Specification of the Contingent Market
journal contributionposted on 08.06.2017, 00:35 authored by Smith, Richard, Olsen, Jan Abel, Harris, Anthony H.
Cost-effectiveness, and cost-utility, analyses have historically been the most widely used techniques of economic evaluation applied to the evaluation of health care programs. However, in recent years there has been renewed interest in the use of cost-benefit analysis, which requires the assessment of programme benefits in monetary terms. The emerging consensus is that such monetary valuation is most appropriately obtained using a survey of individual `willingness-to-pay' (WTP) for the program of interest. There are obviously a considerable number of methodological issues and potential biases to be considered in performing such a survey, which may be grouped into three main areas: (i) the construction and specification of the contingent market; (ii) the administration of the survey; and (iii) the analysis and interpretation of the WTP data. In addition, there are a few issues which also warrant consideration, such as assessing validity and reliability and the impact of ability to pay and income distribution issues. This paper is concerned with assessing the construction and specification of the contingent market. The other two areas being considered in companion Working Papers. This contingent market, fundamentally, consists of four elements: (i) a description of the (benefits of the) commodity to be valued (the `scenario'); (ii) the means by which the WTP value will be administered (payment vehicle); (iii) whether the commodity is valued under conditions of certainty or risk; and (iv) the time period for valuation. Each of these elements of the contingent market is dealt with in this paper. In addition, there are two important biases which may occur as a result of the contingent market specification: (i) hypothetical bias, where WTP responses to the exercise do not represent actual valuations; and (ii) strategic bias, where there are elements within the exercise which encourage `gaming' of the exercise. Again, these two are discussed in this paper. The paper considers the literature relating to the issues mentioned, and uses this to derive a set of `recommendations' for current `state of the art' conduct of WTP surveys with respect to market construction. WTP studies conducted to date in health and health care are then reviewed with such recommendations in mind, to assess the degree to which they reflect this `state of the art'. It is concluded that such studies perform poorly when judged in this manner.