Despite Olga Kryshtanovskaya’s disapproval, I thought it would be interesting and useful to compile a comprehensive list of blogger, pundit and “expert” opinions on the extent of fraud in the 2011 Duma elections. Interspersed among these opinions and analyses are results from federal opinion polls and other evidence.
In general, it seems we can identify three “theses” or “clubs.” The 0% Club holds the idea that falsifications were non-existent or minimal; it is advanced by Kremlin officials and supported by many opinion polls. Its polar opposite is the 15% Club, which is supported by several statistical analyses; its adherents include the liberal and non-systemic opposition. The 5% Club argues that United Russia should not have gotten a Duma majority, but many of their proponents believe that the elections are legitimate nonetheless. Estimates range from 2% to 10%, with a wealth of opinion polling and statistical analysis in support. Most of the systemic opposition and arguably most Russians belong to this club.
The 0% Club (<2% fraud)
* PRE-ELECTIONS POLLS: Levada (53%), VCIOM (53.7%), and ISI (49.6%) all gave United Russia more than its official result. However, because of the relative passivity of United Russia’s electorate – its supporters are less like to vote – raw polling numbers almost certainly overstate its real results.
* OFFICIAL RESULT: United Russia has 49.32% according to the Central Elections Commission of the Russian Federation. This gives it a mandate of 238 seats in the Duma.
* OFFICIAL OPINION: Vladimir Churov, the Chairman of the CEC, has argued fraud is minimal in Russian elections and considers these to be the best conducted elections in the past 20 years. His arguments are based on a suspiciously small sample of just 25 regions where there is reason to suspect fraud was low.
* EXIT POLL: State pollster VCIOM gave United Russia 48.5%, within its 2% margin of error. They covered 62 regions, 1764 stations, and 250,000 voters.
* POST-ELECTIONS POLL: Independent pollster Levada reports 48% responding they voted for United Russia, within its 3.4% margin of error.
* Mark Sleboda, a Eurasianist thinker, thinks fraud was at about 2% but benefited all parties equally.
* OFFICIAL OPINION: Dmitry Peskov, Putin’s spokesman: “Even if you add up all this so-called evidence, it accounts for just over 0.5% of the total number of votes.”
The 5% Club (2%-10% fraud)
* LOSS OF DUMA MAJORITY: United Russia falls below 225 seats in the Duma once its result dips below 46.4%, meaning that fraud of 3% or more starts having a significant impact on the political balance.
* Gordon Hahn, a Russia analyst estimates 3% fraud: “In reality [United Russia got] 46%. More distortion from media control and pressured and voluntary collective voting in the nationally, especially titular Mulsim republics.”
* Andrei Liakhov, lawyer and Russia politics watcher, thinks 3%-4%.
* Patrick Armstrong, a Russia analyst estimates “less than 5%” because “opinion polls rule!”
* @grafomonka thinks it’s around 5% thanks to “fraud in Moscow & ethnic republics.”
* CROWD WISDOM?: In a post-elections Levada poll, a simple majority of Muscovites believed that either fraud was minor and would only have knocked back United Russia by a couple of percentage points or that it was serious and should have lost United Russia its majority. So their average viewpoint on fraud at the federal level seems to be around 5%.
* Readers of this blog, on average, also support the 5% Thesis.
* The KPRF gave 5.4% less to United Russia in a parallel count of CEC protocols during the 2007 Duma elections; may be of relevance as vote rigging appears to have been comparable between now and then.
* Mark Galeotti, a Russian crime analyst, estimates “maybe 5% but not evenly distributed”, with the caveat that it doesn’t include possible cases of coerced voting.
* Andy Young, a Russia blogger, estimates at least 2.5% from the Caucasus alone, so I’m grouping him with Toth-Czifra whom he inspired.
* Andras Toth-Czifra, a Russia blogger, estimates 5-6% fraud using Young’s method for the entire country.
* Mark Adomanis, a “Russia hand”, guesses 5-6% with “huge regional variations.”
* hist_kai, a programmer blogger, estimates 5-7% fraud using statistical analysis.
* EXPERT ANALYSIS: Sergey Zhuravlev estimates 5-6% fraud using statistical analysis. This approach has methodological flaws.
* THIS BLOG’S AUTHOR: Anatoly Karlin estimates “the aggregate level of falsifications is probably at around 5%, and almost certainly less than ten per cent.”
* Nils van der Vegte, Russia blogger, says “I would support ur claims master :P.”
* EXIT POLL: State pollster FOM gave United Russia 43.1%, implying possible fraud of 6.2%. They covered 80 regions, 800 stations, and 80,000 voters.
* EXPERT ANALYSIS: A study by Samarcand Analytics (Alex Mellnik, John Mellnik and Nikolay Zhelev) estimates 6.6% fraud using statistical analysis.
* Joera Mulders, a “Russia watcher”, argues that the “questionable percentage of votes received by United Russia is about 5%-10%.”
* William Partlett, Russia analyst, estimates 5%-10%.
* @biznesslanch makes a “reasonable guesstimate” of 5%-10% in “most places.”
* Juha Savolainen guesses 7%.
* EXPERT OPINION: Aleksandr Kireev, a prominent Russia elections analyst, estimates 8-9% fraud. He also built a map of fraud estimates by region.
* Ani Wandaryan’s “wild guess” is that there was 8%-10%, not including coerced voting.
* LOSS OF LEGITIMACY: Once fraud begins to exceed 10%, is it fair to say that the Duma ceases reflecting the will of its electorate?
The 15% Club (10%+ fraud)
* Max Sjöblom, a Russia blogger, thinks “about 10%.”
* Sean Guillory, a Russia blogger, thinks 10% fraud (or that it is “closer to” 10%-15% than to 5%).
* Nina Ivanovovna, a Russia blogger, thinks 10% or “a little more.”
* EXPERT ANALYSIS: Dmitry Kobak, a programmer, estimates 11% fraud using statistical analysis. But his key assumptions are questionable. He also built a map of fraud estimates by region.
* Alexey Sidorenko, Runet analyst, thinks “between 10-13 mln votes” were stolen, which translates to 9%-13% fraud.
* EXIT POLL: ISI gave United Russia 38.1%, implying possibly fraud of 11.2% but it is constrained by a low sample. They covered 24 regions, 81 stations, and 2562 voters.
* BACK TO 2003: Fraud of greater than 12% would mean that United Russia should get 37% or less and hence fewer Duma seats than in 2003.
* NON-SYSTEMIC OPPOSITION LEADER: Boris Nemtsov claims 13 million votes were stolen, or about 12.5% fraud.
* MACHINE DISCREPANCY: The difference between machine ballots and the overall result for United Russia is 12.8% as calculated by Sergey Shpilkin, only counting regions that have machines. This approach has methodological flaws.
* Gregory White and Rob Barry, writing for the Wall Street Journal, say irregularities “cast doubt” over “as many of 14 million” votes, or about 14% fraud.
* OBSERVERS: The site RuElect, a site that collects election protocols, tallies 34.75% for United Russia (as of 12/28), implying possible fraud of 14.6%. Possible problem: are not observers likelier to send it protocols that don’t match official results?
* EXPERT OPINION: Aleksandr Shen’, a prominent Russia elections analyst, gives United Russia a range of 30%-40%, translation into 9%-19% fraud.
* EXPERT ANALYSIS: Sergey Shpilkin, veteran Russia elections analyst, estimates 15.6% fraud using statistical analysis. But his key assumptions are questionable.
* Fabian Burkhardt agrees with Sergey Shpilkin’s 15.6% estimate.
* Andrei Piontkovsky, an opposition activist writing for the Wall Street Journal, claims 15-20% fraud.
* Grigory Yavlinsky, a Yabloko leader and Presidential candidate, claims 20%-25% fraud.
* Garry Kasparov, an opposition activist, claims fraud of up to 25%.
* MINSK STATION: The discrepancy between pre-elections polls and the official result of the 2010 Belarus Presidential elections suggested 40%-45% fraud. This is what an unambiguously fraudulent election looks like.
* @Pistorasia thinks it’s from 80%-99%.
Fraud In Moscow
As Moscow is generally suspected to have experienced greater fraud than the federal average, and has hosted most of Russia’s protesters, it would be appropriate to create a separate section for the capital.
* OFFICIAL RESULT: United Russia has 46.6% according to the Central Elections Commission of the Russian Federation.
* EXPERT ANALYSIS: Aleksandr Zhuravlev estimates minimal fraud in Moscow using statistical analysis. This approach has methodological flaws.
* OBSERVERS: Examination of observer protocols from stations serving 10% of Moscow’s electorate reveals a 3% discrepancy between the official results.
* CROWD WISDOM?: In a post-elections Levada poll, on average Muscovites estimated that United Russia’s real score was 35% in the capital, implying possible fraud of 12%. By party affiliation, these estimates were: United Russia – 45%; Fair Russia – 34%; KPRF and LDPR – 30%; Yabloko – 26%.
* OBSERVERS: The site RuElect, a site that collects election protocols, tallies 33.04% for United Russia (as of 12/28), implying possible fraud of 13.6%.
* POST-ELECTIONS POLL: In a post-elections Levada poll 32% Muscovites said they voted for United Russia (with a 4.3% margin of error), implying possible fraud of 15%.
* BACK IN 2009. The difference between the average of two post-elections Levada polls (46.1%; 54.5%) and United Russia’s official tally (66.3%) in the 2009 Moscow Duma elections was 16.0%. But also note that in a Levada pre-elections poll 59.3% said they intended to vote for United Russia, a difference of 7.0%. That said, it’s worth noting that Moscow communists report the numbers of complaints were an “order of magnitude” less now than in the 2009 Moscow Duma elections.
* OBSERVERS: The “Citizen Observer” initiative points out Moscow polling stations where no major irregularities were observed reported 30.3% for United Russia, implying possible fraud of 16.3%. Stations where no irregularities at all were observed reported 23.4% for United Russia, implying possible fraud of 23.2%. Criteria by which stations were chosen to be monitored not stated.
* MACHINE DISCREPANCY: The difference between machine and hand ballots for United Russia is 16.6% as calculated by Aleksandr Kireev, by comparing regions with machine voting and those without. Similar results are obtained by Maxim Pshenichnikov, Dmitry Kobak, and Dmitry Oreshkin.
* EXIT POLL: ISI gave United Russia 27.6%, implying possible fraud of 19%. But as noted above, it is constrained by a low sample; note they gave 49.3% for United Russia in St.-Petersburg, where its real result was 33.5%.
* EXIT POLL: State pollster FOM gave United Russia 23.5%, implying possible fraud of 23.1%.
The Moscow Protests
How many people turned up to the Meetings For Fair Elections at Bolotnaya (Dec 10) and Prospekt Sakharova (Dec 24)? Police estimates converge around 20k-30k; the organizers tend to throw up figures from 120k-200k. As both have a dog in the fight, I prefer to trust the geodesic engineer Nikolai Pomeshchenko, who estimated 60,000 at Bolotnaya and 56,000+ at Prospekt Sakharova. As he acknowledges the latter to be an understatement, let’s assume it to be perhaps 80,000. (Novaya Gazeta claims at least 102,000, but they only counted people going in, not those going out early).
According to a Levada poll of the protesters at Prospekt Sakharova, 56% claimed to have attended the Bolotnaya rally. If this is accurate, attendance at Prospekt Sakharova could not have been massively larger – i.e., by the commonly cited factor of two – than attendance at Bolotnaya (obviously, not everyone who went to Bolotnaya was able to or willing to go to Prospekt Sakharova). This implies that it is mostly the same people attending the protests. I suspect that with two weeks of preparation and advertising, the original Bolotnaya “hard core” had the opportunity to agitate some of their social network friends into going. That attendance was only marginally higher at the second Meeting would appear to indicate that “revolutionary momentum” is not building up. Navalny promised one million protesters for the third Meeting next February, so we’ll wait and see.
UPDATE: According to Pomeshchenko, there were 62,000 at the Bolotnaya II meeting (Feb 4), which is also in line with opinion polls. He also estimated 80,000 at the Anti-Orange Meeting at Poklonnaya (Feb 4). RIA Novosti estimated 53,600 for Bolotnaya and 117,600 for Poklonnaya.
My best articles on election fraud in Russia:
- Truth and falsifications in Russia (Al Jazeera)
- Are Russian Elections Rigged?: Opinion Polls Speak Louder Than Western Rhetoric (Sublime Oblivion)
- Measuring Churov’s Beard: The Mathematics Of Russian Election Fraud (Sublime Oblivion)
One of the most comprehensive summaries of the statistical evidence for Russian election fraud is Выборы и статистика: казус «Единой России» (2009, 2011) by Aleksandr Shen’.