> The analogy doesn't hold because in physics hard rules exist. Universality. In sociology you're dealing with a chaotic system where there can be no exact law.

Can anyone talk about cause and effect in sociology (as well as other soft disciplines)? Yes,or no?

If no, then it is pointless to attempt to talk rationally about anything we do in a sociological context having any kind of effect. All policies would be equally (in)effective, so no policy can be preferred to another on account of outcome.

If yes, the analogy still holds. While such contexts are more complex, and do have many hidden variables, unless and until those hidden variables are exposed so as to demonstrate why the old effect would arise from some novel cause, it is rational to continue to presume that the earlier cause and effect relationship is still accurate. That's Occam's razor. Efforts to presume a whole new regime of cause and effect are claims which require data and argument. Without such, they are crackpot theories, often motivated by an ideological agenda independent of reality.

> If only the New Deal was actually responsible for prosperity... ;)

Prosperity for WHO, is the relevant question.