Following the estimates of other Russia-watchers and the call on Twitter from Anatoly Karlin, I joined the guessing game about the amount of fraudulent votes at the 4 December Duma elections. It is worth noting that according to Dmitry Peskov, the spokesman of Vladimir Putin, irregularities accounted for a maximum of 0.5% of total votes, while opposition activists mostly put the figure to 10-15%. The general rough assumption on Twitter seemed to be around 5%. This compilation perfectly shows off the fact of rigging, and points at the most problematic regions, but gives no palpable estimates about the amount of forged votes. When making my spreadsheet, I used similar tools to those used by Andy Young on Siberian Light, but I also tried to draw more consequences from it. Here are the results.
First, the facts. Officially, the turnout of the election stood at 60.21% with United Russia getting 49.32% of the votes (data about the official results drawn from the Central Electoral Commission’s webpage). Turnout showed enormous differences between regions and so did rigging, according to my assumption. I considered as the most problematic regions those where United Russia’s share surpassed 1,5 times the national average (49.32 x 1,5 = 73.98%). These regions were: Dagestan (91.44%), Chechnya (99.48%), Ingushetia (90.96%), Kabardino-Balkaria (81.91%), Karachay-Circassia (89.84%), Mordovia (91.62%) and Tuva (85.29%). It is worth noting that in these regions the vote share of UR also surpassed 80%. Another common feature for the said seven regions was an inexplicably high turnout, above 85% everywhere, which produced the skew in Anton Nikolenko’s distribution curve. I took these seven regions for “cheaters”.
Following Andy Young’s method, I pretended these regions to be average, with a turnout of 60.21% and a vote share for UR of 49.32%. This produced a difference of 2,285,635 votes against the official results, which, after a similar correction in the national turnout, brings UR’s vote share down to 47.63%. If we take into consideration the last opinion poll of VTsIOM before the vote, published on 29 November, days before the election, where UR’s rating stood at 41%, we get a slightly larger number, 2,492,024 forged votes, bringing down UR’s total vote share to 47.31%.
So far so good. I decided to include another step, though, looking at another region where there was an overwhelming evidence for large-scale fraud: Moscow. Here I set the official results of UR (46.62%) against the exit poll commissioned by FOM, in which the party stood at a mere 27.5%. This gives us another 1,161,273 fraudulent votes (assuming that all other parties together got 72.5% of the real votes), bringing down UR’s total vote share to 45.8%, or 45.47%, if we take into consideration the VTsIOM poll (see above).
All in all, in the aforementioned seven regions and Moscow, the total number of fraudulent votes stands at 3,446,907 or 3,653,298, and accounts for 5.25% or 5.56% of total votes.
Taken into consideration the apparent flaws of the method (it assumes an average turnout where a higher-than-average (but certainly not this high) turnout may have occurred, but arbitrarily does not consider any more of the 83 regions of Russia, where further fraud may have occurred), these results, I think, give a fairly strong proof for my estimates about the total proportion of fraudulent votes standing at 5-6%.
Spreadsheet can be downloaded from here.