he suggests that the neo-liberals [some of them like to call themselves neo-keynesians nowadays; that's what the textbook i learned out of labelled itself] pointed to classical theory as the only scientific approach. but, if you've ever taken a look at these linear arguments with these intersecting curves...they're often naive to the point that they defy basic logic. my background is in logic; i didn't take an introductory course in economics until i got to grad school. i remember trying to tackle these arguments from a purely rational standpoint and coming to the conclusion that the models were either tautologies or useless - the assumptions required to make them hold were essentially equivalent to already knowing the outcome. i concluded that the supposedly beautiful simplicity of classical economics really ought to be a labyrinthine theory of complex suppositions and lengthy chains of confusing corollaries.
but, out of that, i think the more profound conclusion i came to is that the whole concept is really madness and should be abandoned.
humans are not electrons. the whole premise that there might be something like economic laws is rooted in the homo economicus model - which assumes perfect knowledge and perfect clarity. that much has been tackled by the behaviourists, who i think are the economic thinkers of the future, but there's really a more basic problem. it's rooted in the assumption that people even care. that was the part i ended up struggling the hardest with trying to rationalize. do people really care that they get $5 instead of $6? some people do, clearly. i'd suggest that most people probably don't. most people would probably rather spend time with their kids, or maybe have a beer with their friends, then sit around and scheme about the optimal outcome. much has been written on this protestant work ethic. in truth, i think assuming that it exists at all is a rather hefty error. and, if you come to this conclusion that people might not care what the optimal outcomes are (or measure optimal outcomes in entirely different ways) then they become entirely unpredictable - even using behaviouralist models.
if i have a desire to maximize artistic output and you have a desire to maximize financial gain then we're going to behave very differently. i might choose social assistance over a 500,000/yr job because free time is more valuable to me. there's no existing model that can make sense of that; it's total anarchism. yet, the possible motives grow very quickly when you sit down and try and enumerate them; there may be as many motives as there are people.
it breaks the causality in the models. because we're not electrons; we're individuals.
and, what a reasonable economic system needs to base itself on is the realization that you can't really aggregate human desire, and by extension you can't aggregate human behaviour through incentive systems and linear causality. you need to look at things carefully, on a case-by-case basis. you need to consult data. you need to listen to what people are saying.
economics is not a science.