Background: The credibility of contingent valuation studies has been questioned because of the potential occurrence of scope insensitivity, i.e. that respondents do not react to higher quantities or qualities of a good. Objective: The aim of this study was to examine the extent of scope insensitivity and to assess the relevance of potential explanations that may help to shed light on how to appropriately handle this problem in contingent valuation studies. Methods: We surveyed a sample of 2004 men invited for cardiovascular disease screening. Each respondent had three contingent valuation tasks from which their sensitivity to larger risk reductions (test 1) and to change in travel costs associated with participation (test 2) could be assessed. Participants were surveyed while waiting for their screening session. Non-participants were surveyed by postal questionnaire. Results: The sample was overall found to be sensitive to scope, testing at the conventional sample-mean level. At the individual respondent level, however, more than half of the respondents failed the tests. Potential determinants for failing the tests were examined in alternative regression models but few consistent relationships were identified. One exception was the influence of more detailed information, which was positively associated with willingness to pay and negatively associated with scope sensitivity. Conclusion: Possible explanations for scope insensitivity are discussed; if cognitive limitations, emotional load and mental budgeting explain scope insensitivity there are grounds for rejecting valuations, whereas other factors such as the alternative theoretical framework of regret theory may render insensitivity to scope a result of rational thinking. It is concluded that future contingent valuation studies should focus more on extracting the underlying motives for the stated preferences in order to appropriately deal with responses that are seemingly irrational, and which may lead to imprecise welfare estimates.
Applied Health Economics and Health Policy, 2012, Vol 10, Issue 6, p. 397-405