The paper subtly slips from "correlation" in the 2nd and 3rd paragraphs to "impact" in the 4th. Are you arguing causality? Are the DB indicators a better predictor than per capita GDP? If so, please link to the paper so people can assess your argument. You have a nice, upward-sloping line in Figure 1, but without any further information we don't know the explanatory value of the data. I'm not saying that the hypothesis is wrong, but the presentation in this blog post fails Econometrics 101.