Comment by int_19h

They do, actually, if you choose a consistent set of basic axioms. Just as in math.

Problem is, unlike math, there's no particular objective reason to prefer some axioms over the others, and so there's substantial disagreement over which ones are "correct". And, of course, depending on which ones you choose, the conclusions derived from them can be radically different, opposite even.

But this is still formalizable - you can make statements such as "from an utilitarian perspective, X is the preferred course of action". You don't have to agree with that perspective for the conclusion to be testable and practically useful.