Abstract
Although contingent valuation is the dominant technique for the valuation of public projects, especially in the environmental sector, the high costs of contingent valuation surveys prevent the use of this method for the assessment of relatively small projects. The reason for this cost problem is that typically only contingent valuation studies which are based on face-to-face interviews are accepted as leading to valid results. Particularly in countries with high wages, face-to-face surveys are extremely costly considering that for a valid contingent valuation study a minimum of 1000 completed face-to-face interviews is required. This paper tries a rehabilitation of mail surveys as low-budget substitutes for costly face-to-face surveys. Based on an empirical contingent valuation study in Northern Thailand, it is shown that the validity of mail surveys can be improved significantly if so-called Citizen Expert Groups are employed for a thorough survey design.
Acknowledgments
The authors are grateful to two anonymous reviewers whose thorough and detailed comments were extremely helpful. The empirical survey of this study was conducted within the framework of the Collaborative Research Centre ‘Sustainable land use and rural development in mountainous regions of Southeast Asia’ funded by the German Science Foundation (DFG).
Notes
1. Apart from procedural arguments against mail surveys, e.g. non-representative sampling, non-response, lack of control of question order etc., the NOAA panel, however, does not provide empirical evidence to substantiate this recommendation.
2. As Diener et al. (1998) has already shown, more than half the CVM studies in the context of health economics are conducted as mail surveys.
3. ‘Yea-saying’ differs from starting-point bias in that responses to the first bid are already biased upwards. Boyle et al. (1998, p. 62) argue that this effect stems from respondents' perceptions that the bid conveys information on the quality of the proposed public good. Consequently, ‘yea-saying’ violates the prerequisite that all respondents should value the same good in a CVM study.