The calculations undertaken in this model are only an approximation of the actual income and expenditure patterns of households partaking in the survey.
Many would defend utility theory (as well as other aspects of their standard model of behaviour) as a useful approximation, pointing out that exceptions to approximations can always be found.