Richard Miller also condemns a form of reductionism he calls methodological individualism. This is the doctrine that all group behavior can be reduced to the behavior of individuals. In one sense, I think individualist reductionism has to be right; surely groups can only do things via individuals doing things. This does not, however, mean that there is no value in studying groups as such, and Miller thinks methodological individualism involves a stronger form of reductionism which not only says group talk can be reduced, but largely advocated abolishing group talk.
One of Miller's examples of what's wrong with methodological individualism involves attitudes toward slavery in the American South. He claims that around 1820, there was a dramatic change in such attitudes. Prior to that point, there was a widespread feeling that slavery was a problematic institution, perhaps necessary for now but something that would eventually die out. From 1820 on, the southerners became much more inclined to think that slavery was actually a good thing, because it exerted a civilizing influence on the blacks.
Miller argues that this shift occurred because of technological changes; the cotton gin made a form of slavery-based agriculture extremely profitable. However, of course none of the southern slave owners would have admitted that this was their motive, and Miller suggests that there is no reason to think they were being disingenuous in giving other motives. Miller concludes that understanding what happened requires looking at the interests of the slave-owning class, rather than focusing on the motives of individual actors.
I find this particular argument stunningly weak, all the more annoying for the fact that I happen to agree with the conclusion. Economists, surely the worst offenders when it comes to methodological individualism, do not in any way require that the self-interested motives of the agents they hypothesize be consciously pursued. They don't care about psychology. And analysis in terms of the self-interest of agents, with self-interest largely inferred on the basis of what people actually pursue, has a good record of success in providing economic explanations. It is not at all obvious that this pattern of individualist explanation is not adequate to the case Miller describes. You're welcome to read his actual work, of course, but I don't recall him including any additional detail in the case which would block this response.
Nonetheless, there are clearly anti-individualist theories (not necessarily non-reductivist, but theories that insist on examining groups as groups). No doubt they are the kinds of theories Miller is interested in. It is overwhelmingly plausible that the upper classes have class consciousness, and always have had. Plato thought so. Marx (obviously) thought so. Nietzsche thought so. Adam Smith even thought so. Pretty much every historian I'm aware of thought so. The explicit writings of the upper classes themselves support the hypothesis that some kind of group loyalty is an important motive in the behavior of members of the upper classes. Really anybody who isn't a political libertarian seems to have noticed this.
It is also arguable that the lower classes have often had class consciousness, and of course Marxists of Miller's type would tend to argue that part of the reason modern capitalist institutions get away with so much inequality is because they do a good job of undermining the class consciousness of the lower classes, fragmenting them into mutually hostile groups.
It is quite plausible that any group which is successful and powerful must have mechanisms for encouraging group loyalty, various rewards for supporting the club and punishments for betrayals of the group. While investigating exactly what those mechanisms are is, of course, a fascinating project in itself, for an old, well-entrenched group, they're likely to be quite diverse and complicated. Thus, for some purposes it may be more useful to just study the group, noting that it engages in purposeful behavior and that the interests of the group sometimes explain the behavior of individuals more perspicuously than trying to tease out all the little incentives that cause a particular individual to act in the group's interest in a particular case.
This pattern of explanation again particularly makes sense for powerful groups that are effective at perpetuating themselves, and so it seems natural that this form of explanation should work better for ruling classes than for lower classes (which seems to be the pattern in historical explanations). Still, whether this sort of group-based approach to explanation is overall desirable is clearly an empirical question. I'm inclined to think this form of explanation probably is desirable, but Miller's example, and his discussion of it, are woefully inadequate to establishing that fact. Much more wide-reaching investigation, a much greater variety of examples, and ideally examination of a variety of disciplines would be needed to establish that claim.
Which, incidentally, brings up a quick bleg. Anybody know of any good recent work that's been done in this area? I'm sure somebody must have a better discussion of these issues than Miller.