Did you read any of the actual studies cited? The modus operandi in social science is to throw in enough variables into your regression (or leave out a few key ones) until you get the result you want. Just throwing a lot of "[Nunn et al 2007] say x " on a page is not evidence. If there is a particular one of those papers that you thought was the most convincing, I can take a look at it and evaluate the actual evidence/data presented.