Neoclassical economics is the idea
that humans are rational, utility-maximizing agents with fixed preferences, that they make decisions ‘at the margins’ and that the mechanisms of supply and demand (operating free of government interference) will lead to a general equilibrium whereby resources are allocated efficiently
according to this article in The Nation.
I’m not sure this is right. Neoclassical economics articulates *assumptions* precisely to make certain predictions about human behavior. Historically, those assumptions have been rationality, complete markets and prices taking. The predictions generated from those assumptions can be tested. They are often verified (quantity demanded generally goes down when prices increase, for example) and when they’re not they provide a way of talking about negative results.
David Card famously found that minimum wages in New Jersey didn’t decrease employment in the fast food industry. The neoclassical framework gives us a means of discussing that result; its a sounding board. Why didn’t demand for low skill labor slope downwards? Was it because the assumption of rationality, price taking or one of the other assumptions of the theory was violated? Were prices (in this case wages) being measured correctly*?
That Card discovered data that contradicted the theory isn’t enough. We need to know *why* the theory failed. Its not enough to scream from the sidelines “the theory is wrong!” To make progress, we have to correct the theory by updating assumptions.
The neoclassical theory has traction because it has plausible assumptions, it implies tractable predictions and it has been useful. First, it seems likely that people use reason when making decisions. Obviously, that’s not strictly true and its not true that people are good at reasoning, but this is a feature of human behavior. It makes sense to try understand the implications of that feature. Also, unless you talk of certain industries, or even particular firms, the price taking assumption is the only general assumption one can make that is approximately true**.
Second, you can write out explicit models of neoclassical theory. There’s equations. They show precisely the relationships the modeler is interested in, no more and no less.
Third, neoclassical theory has delivered the goods in terms of scientific results (of which, I include Card’s).
Like any science, economics advances by identifying, precisely, the flaws in the theory and then updating the theory. More from the article:
some mainstream economists dismiss heterodox work as quackery, others claim that the mainstream has actually assimilated many of the heterodox critiques. (You’ll note that these two responses, both fairly common, are also logically incompatible.)
…but this isn’t incompatible. Real negative results from “heterodox” economics get incorporated***. Much of the inspiration for behavioral economics is from the “finding” that people aren’t rational, for example. I often feel that “heterodox” economists want to do theory-free economics. Theory-free economics is quackery.
UPDATE: Rodrik says is better: “To me [neoclassical economics] represents nothing other than a methodological predilection for deriving aggregate social phenomena from individual behavior–and as such it is a very useful discipline for any social science. You say people have some preferences, they face certain constraints, take others’ actions into account, and go from there. Neoclassical economics teaches you how to think, not what to think.”
UPDATE 2: HedgeFundGuy talks heterodox: “Like any science, there are those who think the current mainstream is really wrong about big issues. But these overbroad criticisms are lame because they offer nothing specific as an alternative, just a critique of the imperfect status quo. There’s mention of recognizing customs, norms, etc., but that’s just sociology, and that ‘science’ has hardly been more successful than economics… They have never made a statue in honor of a critic.”
UPDATE 3: I think Borjas picks up the grain of truth in the Nation article. Economists are herd animals.
*I’m thinking of non-wage compensation.
** I’m not an IO expert, but there’s certain competitive structures where industries with only two firms act like price takers (i.e. not like monopolies).
*** In the article, Card’s results are given as an example of a heterodox result.