I'm late to the discussion (as usual); but this seems precisely the kind of thing that should have a 'correlation =/= causation' warning flag on it.
If you want to toss out smart sounding rhetoric in a simplistic fashion with only minimal relevance, sure. While correlation does not prove causation, the cause *is* going to be present within the set of correlative data, right (assuming a complete data set)?
If I walk into a room and flip a switch on the wall and the light turns on in the room, it's reasonable to look at the set of things that happened between when the light wasn't on to when it was. I walked into the room and I flipped a switch. The light could be on a sensor which detected me walking into the room and the switch did nothing, or the switch could have turned on the light. But assuming I know that those are the only two changes which occurred, one of them must have been responsible for the change of state in the light.
It would be absurd to instead embark upon a statistical analysis of how often the lights in any given room in the house are on or off over time, but that's essentially what Allegory was trying to argue we should do to figure out why the light turned on. It's ridiculously convoluted thinking and makes no sense at all. It's the kind of thing one might suggest, not to clarify the question at hand, but to muddy the issue with extraneous and useless data (which I suspect is exactly why he brought it up).
If you want to know why the GOP lost in 2008, you look at the set of things which changed between then and 2004, when they won. That is the set of possible causative factors. Then you assess each factor to determine to what degree each may have had an impact on the resulting change. This is how anyone who understands critical thinking goes about doing things. Surely, you agree?
And that's what I did. Not sure what Allegory was trying to do though.