There is a similar explanatory role, Railton thinks, for morality (the "comprehensive" criteria given by the impartial "moral point of view" -- roughly, preference utilitarianism). We may think it is not a coincidence that we have abolished slavery. Abolitionism is partly explained by the fact that slavery is unjust. You may suggest that we can just as well appeal to the mere beliefs of the abolitionists. But Railton points out that even unrecognized injustice may be conducive to social unrest. The mere fact that the interests of some group are being systematically discounted will tend to foster dissatisfaction among those group members, even if they don't consciously appreciate what's going on.
An individual whose wants do not reflect his interests or who fails to be instrumentally rational may, I argued, experience feedback of a kind that promotes learning about his good and development of more rational strategies. Similarly, the discontent produced by departures from social rationality may produce feedback that, at a social level, promotes the development of norms that better approximate social rationality. (p.193)
Hence Railton predicts that "over time, and in some circumstances more than others, we should expect pressure to be exerted on behalf of practices that more adequately satisfy a criterion of rationality." (pp.196-7) Justice is thus seen as a selection pressure in cultural evolution.
But, as others noted in class, this explanatory role fits ill with Railton's utilitarianism. Some form of egalitarianism - which places greater weight on the 'separateness of persons' - would seem to better predict social stability (given human selfishness). That's no reason to favour egalitarianism as a normative theory; but it does raise the question of what distinctive explanatory role a utilitarian moral theory is thought to have.
I prefer a different route. We shouldn't expect normative facts to have any direct impact on society. But they'd better influence us as rational agents!
Often (we may think), we hold the beliefs we do precisely because they are the most reasonable ones to hold. So if the normative fact of slavery's immorality simply consists in this being the most reasonable view, then the normative fact itself explains why we believe it (given our general capacity for reasonable belief). And that belief in turn influences our actions.
This account takes us far away from the kind of naturalism that would see ethics as just another branch of scientific inquiry (sociology, say). But I think that is a good move.