Comments and Questions

The paper proposes to test experimentally the cardinal utility maximization hypothesis, and concludes that the experimental data supports this hypothesis. There is data from two variants of a choice experiment where subjects buy quantities of a few different goods, three kinds of nut in one case, and apples, pens and facial tissues in the other case, plus a variant where the subjects bid on the goods. Although the experiments seem quite basic the exact design is not easy to understand and instructions are not provided. Only a minority of the subjects seem to have completed the experiments (on page 12 it says that only 38 of 105 subjects delivered valid data in Choice I etc.), which does suggest that there might have been some problems with the design. Thus it is not clear what information might be gained from this experimental data. In general, I find it difficult to see how an experiment of this kind could be used to gain information about cardinal utility. I’m not sure what to suggest, perhaps to look at the paper by Abdellaouia, Barrios and Wakker (2007) which is certainly relevant for the issue.