However, what is important for our research design is this increase in banking was not accompanied by a systematic change in other plausibly exogenous variables such as percent urban or population, which is consistent with the assumption of our research design. It also appears that the divergence in banking did not cause a divergence in real income per capita growth. Specifically, while real per capita income went up by 96% on average in states with below-median spetsbank concentration, it went up by only 74% in states with above-median spetsbank concentration. These patterns are also apparent from Figs. 1 and 2, where Fig. 1 shows the positive relationship between the log of per capita lending during 2002–2006 and the number of spetsbanks in 1995, and Fig. 2 shows the lack of such a relationship between the annualized increase in real income per capita from 1996 to 2007 and the number of spetsbanks in 1995. Along similar lines, Fig. 3 shows that per capita lending in high-spetsbank regions quickly outpaced lending in lowspetsbank regions between 1997 and 2007, while Fig. 4 shows that per capita real income did not.