----------------- HES POSTING ----------------- 
I have found  Dutch economist Bernhard van Praag's work (a.k.a. the Leyden 
School)  in the field of utility measurement rather convincing. In short 
(and incomplete): 
 
His questionnaire approach is based on the assumption that  respondents are 
willing to provide the best information possible. By setting the zero-one 
interval as permissible range, the utility function assumes the 
characteristics of a statistical density function with all its properties. 
The density function with the most embodied information (entropy) is the 
normal distribution. Thus, respondents effectively provide information that 
can be interpreted on the same statistical features as the normal 
distribution. Still, this does not measure cardinal utility but a (series 
of)  level(s)  of   satisfaction "a la" ordinal ranking. Results are not 
comparable among respondents, given the ordinality of responses. 
 
BvP uses the results, i.a.,  to suggest equivalence scales and subjective 
poverty lines. It is a promising approach -- original in its assumption and 
design and rather robust in results. It beats the traditional "political" 
manipulation. 
 
Jesse Vorst 
 
   
 
 
 
------------ FOOTER TO HES POSTING ------------ 
For information, send the message "info HES" to [log in to unmask]