[WS:] I guess it boils down to the meaning of the phrase "relative autonomy." I do not think anyone would seriously argue for "absolute autonomy" but that indeed depends of the historical circumstances. It is possible to think of the Soviet system or even PRI rule in Mexico as a high degree of state autonomy from class interests, but I am having difficulty finding a real historical example of the state being totally subservient to the interests of a single class. Perhaps the "Old republic" of Brazil may be an example of the state ruled by the landed oligarchy, but that ended in 1930 by a 'revolution form above" in which the military acted against the interests of landed oligarchy.
I think the conceptual problem before us is rooted in the fact that classes are rather heterogeneous and so are their interests, whereas organizations are more homogeneous and with more clearly articulated interests. As a result different organizations may have affinity to different articulations of class interests. For example, a labour party (an organization) can represent divergent working class interests - e.g. higher wages demanded by skilled labor unions, or social protections demanded by "unskilled" labor.
Therefore, while this organization represents interests of the working class in some abstract generic - and rather uninteresting in my view sense - the specific historical outcome depends on the interaction between specific organizational goals and its alignment with a specific subset of working class interests. The same can be said of landowning classes, industry owners, and so on.
The key to understand this relationship is that class interests do not express themselves directly, they need organizations to articulate and pursue them, and these organizations may have a varying degree of autonomy from the class interests they purport to represent. That level of autonomy depends, in turn on a whole set of broader socio-political circumstances ranging from socio-demographic composition, to the type of economy, to historical path of development, to international relations and to moves made by competing organizations. This, in my view is the most interesting object of inquiry because it can explain specific historical outcomes.
Generalities like "state or institutions are always linked to class interest" are no different from the platitudes of rat-choice theories claiming that "humans act in their self-interest." They may be true in a very general sense but they tend to fall into a vicious circle or teleological fallacy "whatever happens does so because it was in someone's interests" be it a class or a "rational individual."
This kind of "explanation" is no different than saying "God's will" or some other semantic contraption substituting the feeling of certitude for an explanation. If one assumes that god is omniscient and omnipotent, then such an 'explanation" is the only logical conclusion, because anything happening without god's will would counter the dogma of god's omniscience and omnipotence, so it is apriori ruled false. The same thing happens if we assume that class interests always produce certain outcomes, and claiming that something happened independently of these interests is apriori ruled false because it contradicts the original assumption (or dogma).
The difference between me and Marxists (or perhaps some of them) is that while I do not think that class is typically a sufficient explanation of historical outcomes, I also think that it is possible that in certain circumstances class is the major predictor of outcome and I am perfectly happy to accept this if the evidence supports is. The Marxists, otoh, (or at least some of them) think that the class is major predictor of any outcome, are unwilling to accept a claim that under some circumstances it is not, and dismiss any evidence that supports it.
-- Wojtek
"An anarchist is a neoliberal without money."