It has been suggested that just as in social choice theory it is hard to turn indivudual preferences into a single collective preference, it might be equally hard to turn an individual's multiple wnats and needs into a single individual preference ranking. In other words, if I am hungry, thirsty, cold, bored and tired to various degrees, does it follow that my preferences regarding what to do next, assuming each of my options addresses some nontrivial combination of my wants and needs, have to be consistent?

Sometimes the utility and rational choice model is taken as an action, as in implied preferences, where people decide it is pointless to ask people what they prefer and try to see if their behaviour matches their preferences, and instead postulate that preferences are as they need to be to be consistent with behaviour.

Rational choice theory may be an interesting branch of mathematics, but I think to make it the axiomatic foundation of economics is a bit of a stretch.

Those whom the Gods wish to destroy They first make mad. -- Euripides