The continuing question here is this: how and through what mechanisms do various social entities exercise causal influence with respect to social outcomes? (This will also be relevant to the question: how does power work in a given society?) We think the Gulf Stream wields influence with respect to weather and climate change, but not the circulation numbers of the NYT. And we can specify a lot of detail about the climate-system micro-mechanisms through which this influence is conveyed. We need comparable insight into social causation and various putative social causes or influences.
In that spirit -- do social structures wield causal powers, and through what mechanisms do they do so?
Let's take one specific example -- a mid-size social structure such as the higher-education system found in various countries. This is a complex of institutions, funding arrangements, internal practices and organizational forms. Can the higher-ed system be called a structure? And can it exercise causal influence on the society in which it is embedded?
First, is it a social structure at all (as opposed perhaps to simply an agglomeration of miscellaneous lower-level institutions)? There is no doubt that this system is "heterogeneous" in the sense discussed in prior posts. There are many varieties of public universities, many types and strata of private universities, and many distinctive variations across individual universities. Moreover, these institutions are only somewhat loosely connected together. Nonetheless, British, Russian, French, Mexican, and US universities are different from each other, they have different institutions, values, and philosophies, and they serve in different ways to educate the post-high school populations of their countries. And these institutions have very important distributive consequences in their settings: the credential of baccalaureate gives the young graduate different opportunities than he or she would have otherwise. So it seems justified to describe these systems as "structures".
The causal influence of the higher education system in a given country is fairly easy to address. The features of the higher education system in a country determine the skill level of the educated population and the "culture" of educated people. Schooling systems that favor technical and vocational education will have one sort of influence on economic development, different from a system that favors a broader "general" education. Systems that provide broad and affordable access across social classes have an effect on the inequalities that the society will embody in the future; and these effects are different from those of systems that are class-exclusive. Universities are places where young people learn a lot of their political culture; so a system that embodies a culture of left-ism will have different political consequences than one that embodies quietism or consumerism.
The causal mechanisms of these influences are also easy to discern: effects on individuals (skills and political values) leading to the creation of new economic and political opportunities for society in the next iteration. Other effects are a bit more subtle -- for example, the influence that universities have through creation of a variety of social networks (on Wall Street, in Washington, in the military or intelligence world). In each case the causal powers of the institution are readily disaggregated into the microfoundations through which the institution shapes and constrains the individuals who pass theory the institution.
And we can also reason on first principles (to be tested through comparative research) that different national complexes of institutions, practices, and values in higher education will make for observable differences in those societies' institutions, performance, and collective behavior. This is the central point of the new institutionalism: different institutional complexes doing the same social work will have different outcomes for behavior and the further development of institutions.
So here is one medium-scale example that responds in the affirmative: concrete social structures do possess causal properties; these powers work through the features of constrained agency that they guide; and that we are likely to observe the different workings of these causal properties in different country settings.
(I believe we could work out similar analysis for other social institutions, such as social-property systems, kinship systems, or systems of media ownership and control.)