Proceedings of the German Development Economics Conference, Berlin 2011 33
Are responses to a simple survey item sufficiently reliable in eliciting risk attitudes? Our angle in examining reliability is to conduct comparative research across Thailand and Vietnam. We find, first, that the survey item is informative about individual risk attitude because it is plausibly related to socio-demographic characteristics (including vulnerability), it is experimentally validated and has some predictive power. Second, however, we find major differences between both countries: whereas explained variances of regressions are tentatively higher in Vietnam, the predictive value of the survey item is much lower than in Thailand. Therefore, the survey item cannot be implemented across countries in an unreflected way.
risk attitude socio-economic survey household behavior field experiment