You're asking about the type of causality that would account for the correlation. This article is hardly a detailed theory that accounts for all the variables. Indeed, psychological traits and social positions may be built up in cycles, so both nature and nurture (genetic traits and environments) would be in effect, influencing each other. The aim of providing a full theory of how this would work would be fit for psychologists and sociologists rather than philosophers like me. I'm more of a big picture guy.