This was studied by the Freakonomics guys. The causality goes the other way. Favorite candidates received more money than underdog candidates because donors wanted to back the winner, ie. donors didn't create the winner; rather they jumped on the bandwagon.
would be nice to see how he arrived at that conclusion if you have a link. As is, I'm not sure how one can establish that link w/o it being a self aggravating issue (more spending, more exposure, more popularity). I guess I'm saying is, it sounds like a coupled math problem. I'm not the biggest fan of populists necessarily but I've seen them lose a few races even when they were ahead in the polls when money entered big in favor of their competitors... though one could argue these may be special cases and not the norm.