Economics attempts, a priori, to discover the wants and desires of man, his preferences and motivations. Actions are then logically deduced from these wants and desires. The process is not a posteriori, where wants and desires are logically deduced from actions. In this sense, Economics fails the first empirical requirement of natural science: It is not a posteriori. Also, it fails the second empirical requirement of natural science: It cannot be repeated (meaning, the wants and desires of man cannot be standardized - they are as varied as are people and the circumstances and times in which they find themselves.) Therefore, it must be a social science (and better than any other social science!)
If all wants, desires, preferences and motivations are elusive, it would be impossible to presuppose any axiom in economics. To pressupose an axiom, it is not necessary to know every want, desire, preference or motivation held by each individual person but those which are held in common. These are "ends" that are sought by multitudinous "means," the variety of which is as numerous as the people on earth.
I don't think I said that Human Action makes the case that Economics can be repeated, rather that it cannot and, therefore, fails the test of empiricism in the natural sciences.
Check my blog, if you're a loser
Lyle: "Economics attempts, a priori, to discover the wants and desires of man, his preferences and motivations."