believing the opposite would lower your status in your peer group and it threatens the social order you desire (even if you haven't established that order as a standard for society, it is the standard you believe society should operate by). I don't see the latter as being much more "elaborate" than the former.

Of the two concepts, the former is easier to implement as an automatic, unconscious response, because it doesn't involve multiple levels of abstraction. That's why I'd tend to favor it over the second version.

Even simpler, though, might be to just model certain beliefs as "bad" or "good" and treat those that oppose the "good" as "bad". (However, the controlled variable in that case is still "perceived status", since we learn what beliefs are bad or good through applied status consequences.)