>"But a true correlation in the underlying process always has some causal explanation." In the vast majority of cases, A really is correlated with B to some ex

>"But a true correlation in the underlying process always has some causal explanation." In the vast majority of cases, A really is correlated with B to some extent. People have done this with large social science datasets since at least the 1960s, they find everything is correlated with everything else. So, whether a correlation merely exists is not interesting to begin with (there are some exceptions in physics where theory really predicts zero correlation). Yes, this means all the people currently plugging away at checking for the mere existence of correlations are wasting their time and (probably) taxpayer money.