For those interested in SOCIAL theory, please ignore the other reviews on this page. The fundamental problem with their negative response is a failure to take seriously the central realist concept of emergence (or, for one review, to simply dislike the work because of all the big words; as a philosophy professor of mine once told me, before you criticize philosophy for being 'up in the clouds,' you first have to learn some philosophy).
Archer and other realist's are keen to emphasize that the world contains emergent strata (atomistic, biological, psychological, social, etc.) - i.e. entities which once produced through combinations of more fundamental parts cannot be explained solely by reference to those parts and in fact can act back upon them with causal effect (being raised in a family of racists - a social interaction - can alter ones psychology toward paranoia, fundamental attribution error, etc. - or even effect one's biological state if one decides to go on a hate-filled rant among a group of knife weilding members of the Other).
That's why the agent-structure problem is important, because throwing ones hands up and saying that everything can be reduced to individuals or the dictates of nature have been a dead-end repeated over and again; scientific work based on these premises continually fails to accomplish the goals it sets for itself. So emergence may be hard, and tackling the agent structure problem may seem like an impossible task, but it's the effort that counts so that people aren't led astray by nonsense that soon all will be explained by our 'natural' desire for pleasure, health, procreation or whatever. So to give an example, when the social structure of religion, which defined pleasure as bad unless it worked toward God's will, gives way to the social structure of the media-entertainment industry, our natural drive toward pleasure goes from something like abstaining from sex to seeking it out.