Was Russian election turnout too non-Gaussian to be real?

 We’ve talked about attempts to prove election fraud by mathematical means before.  This time the election in question is in Russia, where angry protesters marched in the streets with placards displaying the normal distribution.  Why?  Because the turnout figures look really weird.  The higher the proportion of the vote Vladimir Putin’s party received in a district, the higher the turnout; almost as if a more ordinary-looking distribution were being overlaid with a thick coating of Putin votes…  Mikhail Simkin in (extremely worth reading pop-stats magazine) Significance argues there’s no statistical reason to doubt that the election results are legit.  Andrew Gelman is not reassured.