We are pleased to see the response by Brazier and McCabe (2007) to our paper (Parkin and Devlin, 2006), as it seems to concur with the spirit in which we wrote it; their response, like our paper, supports a wider research agenda on health state valuation than the current consensus advocates. They also share our view that the apparent consensus with respect to the primacy of standard gamble (SG) and time trade-off (TTO) should be challenged. We agree with their suggestion that emerging health state valuation methods such as discrete choice experiments (DCEs) are a promising route. However, we remain unconvinced by their re-stated case against the Visual Analogue Scale (VAS) for the same reasons that we gave in our original paper: it rests on unsubstantiated and one-sided arguments and assumptions rather than real evidence on the important issues related to this debate.
Brazier and McCabe are in error in what they initially identify as their departure from our views, stating that we limit the term 'stated preference' to monetary valuation. We neither say nor imply that. In discussing stated preference, our aim was to locate the source of the bias amongst economists in favour of choice-based, or indirect, ways to elicit values. We suggested that this is a logical and reasonable requirement in the context of stated preference contingent valuation methods for monetary values, since it is an attempt to mimic the conditions under which markets generate revealed preferences. There is no such economics-based argument relating to the measurement of utilities or health state values, which are not observable within markets.
But in including non-monetary valuation under the label of stated preference, Brazier and McCabe seem to be making their own narrow interpretation of the term 'stated preference,' which they appear to limit to so-called 'choice-based' techniques. This may add to a widespread confusion over the terms revealed and stated preference in the economic evaluation literature in health economics, which is not evident in other areas of application, such as the environment or transport. In the highly influential book Methods for Economic Evaluation of Health Care Programmes by Drummond et al. (2005), it is even claimed that 'choice-based' techniques are revealed preference: in discussing TTO, Paired Comparisons, Equivalence, Person Trade-Off and SG, they say that 'a subject can be asked to choose between two alternatives, thus revealing their preference indirectly . . .. (this) method comes primarily from economics and decision science and is a particular application of the revealed preference approach. (Revealed preference is a general approach in economics whereby the underlying preferences are revealed by the choices that individuals make.