> That is, people are selfish rational actors. The problem is that they arent. Libertarian models of humans are incredibly simplistic.
I can't speak to Friedman, but Hayek's views were more subtle than that (and so are mine, for what little it matters). He recognized many limitations on the rationality of economic actors; he advocated what today would probably be referred to as bounded rationality.
Even bounded rationality is totally wrong. Human behaviour isnt (economically) rational "except for limitations".
The very source of the value we place upon institutions, people, goals, etc. is entirely a product of the interaction between culture and basic psychological drives.
People do not have "preferences" in the strong propositional sense required for modelling. They have drives, ideals, etc. These can be systematized by asking them to rank states of affairs but even here things are delicate (which is why polling on issues is generally useless).
Any kind of "rationality" talk is extremely prescriptive in practice: failing to actually look at the world and describe it, libertarians et al. go around theorizing in an ahistorical vacuum with pseudo-computer human stand-ins. And criticise actions in the real world that dont conform.
Friedman's model does not assume that people are selfish rational actors. It assumes that people have goals and that they tend to take actions they believe will accomplish those goals.
"take actions they believe will accomplish those goals". Well there you go: quite wrong!
People aren't at all clear on what they want, nor how to accomplish it. People respond to impulses and psychological drives not propositionalized goals.
They shifted simplicity from society to the individual. So if the social was complex it was nevertheless based on a the motives of simple individuals.
That is, people are selfish rational actors. The problem is that they arent. Libertarian models of humans are incredibly simplistic.