Neoclassical economists purport to be able to give rigorous and accurate explanations of human behavior. But to explain actions, you have to have an idea of what it is that agents are doing. It used to be that early economists, who were utilitarians, held the view that what we're doing when we act is trying to maximize utility (construed as felt or experienced pleasure). But insofar as pleasure is difficult to quantify and measure, economists decided they needed a more easily quantified view about what agents are doing. Reality must simpler than that, they thought, because otherwise how could it fit into clean formal models? Moreover, some economists bought the argument that social science could be "value free", and they worried that utility (pleasure) was too value-laden and tendentious. At first, economists went with "desire-satisfaction" instead of felt pleasure, but then the problem arises as to how economists can describe what desires people have. The resting consensus today is that economists should just talk in terms of an agent's "preferences" and whether or not such preferences are being "satisfied".
But how do economists know what an agent's preferences are? According to contemporary orthodoxy, an agent "reveals" her preferences in certain behaviors. We look at what she in fact does (e.g. buys a coffee) and infer her preferences from that. It's actually a bit narrower than that: the only behaviors where agents reveal their preferences is in market transactions (and, maybe voting), and the measure of how to rank one's preferences is given by the amount of money they're willing to pay for some thing. When I buy a coffee everyday for a year, economists hold, it is reasonable to say that I have "revealed" a certain preference. It is in this way that economists think that they can explain everything about us that matters.
There are tons of things wrong with this whole setup. I just want to point out one glaring problem.
First, take the problem of environmental degradation. Now, it should be pointed out that neoclassicals are falling all over themselves in trying to explain how their enterprise, which essentially adopts the perspective of the capitalist and spells out what it's "rational" for capitalists to do, can account for the obvious problem of environmental degradation. At first glance, their approach seems to either say that the environment is not of value, or that it only has value insofar as it can be commodified and traded on markets. Contemporary economists are aware that this is a problem. So they are scrambling to try to re-package the same old game in order to fend off some recent criticisms of their discipline.
As is well known, one obvious problem with capitalism is that it encourages investors and capitalists to take a very narrow, short-run perspective. Big profits this quarter are often all that matters. Among it's many problems, this narrow perspective obviously clashes with the long-term perspective we would need to even begin to seriously address the problem of environmental degradation. So what do economists say about this?
Well, according to them, if we care about anything at all, it can be modeled by "revealed preference theory". We all place different emphasis on the value of the future, they'll say, so we all have different preferences regarding the importance of the future (as opposed to the present). How do we know what different preferences people have regarding the future? Look at how they're "revealed" in market activity.
Do capitalists then have a very low preference ranking for the future, and a high preference ranking for short-run profits? Capitalists, it turns out, "reveal" such preferences all the time. And insofar as capitalists are merely satisfying their preferences like the rest of us, they aren't doing anything wrong. Economics, in this way, is "neutral".
But here's the rub. This behaviorist setup is circular. It says that what people want, indeed what it would be rational for them to do, is basically just what they're already doing. That is extremely conservative. It suggests that we couldn't fall short of what we want, that we couldn't fail to be rational. It suggests that what we want just is what we're already doing. On what basis could someone, then, want to change anything? Could I have a "preference" for radically changing society? It seems not, since I'd have to actually change it in order to "reveal" such a preference. This setup is ideological because it suggests that there is no gap between the way we think things should be and the way that they are. It leaves us without very language we'd need to articulate what's wrong with existing states of affairs.
In the case of the narrow-minded capitalist chasing down short-run profits, it seems like the thing we'd want to say is this: she is acting irrationally. By her own lights, she is undermining the possibility of continuing her own enterprise in the long run. More strongly, she's undermining the possibility of doing much of anything at all in the not-so-distant future. But we cannot say this on the "revealed preference" model. For that model says that her preferences regarding the future are simply "revealed" by what she is presently doing. Thus, she's just satisfying her revealed preferences for the present, so there is no self-undermining problem. The theory forces us to say that she is acting rationally after all.