In all of economics that I know, people maximize their Utility $U_i=f(x_1,...,x_n)$, and therefore make some kind of cost-benefit analysis.
One obvious criticism to this model is that we often do not have the right information on the $x_k$ (it is hard to have a parameter for altruism for instance).
But except from that, is this formalism irrefragable ? Or are they theoretical limitations, situations that it cannot account for ?
I would gladly accept any thought or reference on the subject