Citing evidence revolving around pre-elections opinion polls and exit polls, in my Al Jazeera article on the Russian Duma elections I made the argument that “the aggregate level of falsifications is probably at around five per cent, and almost certainly less than ten per cent” (with the caveat that it was far worse in several regions, including – and controversially to some commentators on this blog – Moscow). Since then, the main polling organizations have released a series of post-elections polls, and pre-elections polls that could not be legally published on the eve of the elections that appear to confirm my initial judgment*.
First up was VCIOM, a week ago. We remind that their exit poll predicted 48.5% for United Russia (final result: 49.3%), which is well within the margin of error. Their new polls seem to confirm this. In their Nov 26 poll and their Nov 30 poll, some 37% and 41% said they were going to vote for United Russia, respectively; after the elections, in Dec 10, this had dropped to 35%. However, one has to bear in mind that only 72%, 75%, and 76% respectively displayed a preference for any one party; the rest said they wouldn’t participate, hadn’t decided or couldn’t answer, or intended to spoil their ballots. Adjusting for these, of committed voters some 51% and 55% said they intended to vote for United Russia in the week before the elections on Dec 4 (average 53%, i.e. basically no change from Nov 20)*. The result from Dec 10 indicates a fall to 46% electoral support, but this is after several days of fraud allegations and very negative PR against United Russia and as such not too reliable. The post-elections slide in support for the party of power appears to have been a one-off; as of Dec 17, one percentage point more said they would vote for United Russia.
In Levada’s post-elections poll, some 48% said they voted for United Russia. This closely tallies with the real result of 49.3%, especially considering the 3.4% margin of error. Fewer said they supported the Communists, Fair Russia and Yabloko, 17%, 12%, and 2% respectively, than was the case in real life, at 19.2%, 13.2%, and 3.4% respectively! As with the VCIOM poll, this indicates (but nothing more than that) that the level of fraud may have even been less than the 5%-7% that constitutes my best guess.
But the same could not be said of their Moscow poll**. There, only 32% said they voted for United Russia with an error margin of 4.3% (as opposed to its real result of 46.6%). This supports an avalanche of evidence elsewhere (1, 2) indicating very serious fraud taking place in Moscow (FOM exit poll: 24%; ISI exit poll: 28%; Citizen Observer: 23%; voting machines: 30%). This lends support to my other thesis that one of the key reasons why protests have been especially severe in Moscow (and not in, say, St.-Petersburg, or Vladivostok, hardly bastions of United Russia support either) was because of its quite exceptional level of fraud.
They also asked some direct questions about electoral fraud throughout the country. 56% of Muscovites believe violations were either “minor” or “serious”; 15% believe there were “practically none”, while 14% think there the level of fraud was “absolutely huge” (so average: between “minor” and “serious”). Likewise, 14% believe United Russia shouldn’t have gotten even a third of the seats; 28% think United Russia wouldn’t have gotten a majority in the Duma; 23% think the results would have changed by only a couple or so percentage points; and 15% think they wouldn’t have changed at all (so average: basically 50/50 on whether or not United Russia could have retained its majority). This all tallies very well with my own best guess of 5-7% overall fraud. Asked to estimate the figure that United Russia really got in Moscow, and their averaged answers were: United Russia supporters: 45%; Fair Russia supporters: 34%; the Communists and LDPR supporters: both 30%; Yabloko supporters: 26%. The overall average was 35% for United Russia, which again tallies in with the general weight of evidence about fraud in Moscow. Crowd wisdom!
Contrary to pretty much everyone in the Western media, the impact of the fraud allegations on Putin’s popularity have turned out to be fairly minimal. According to FOM, trust in the Russian Premier went from 54.4% in Nov 27, to 52.8 in Dec 11 and 53.6 in Dec 18. Hardly the end of the world. According to Levada, Putin went from 67% approval (32% disapproval) in November to 64% approval (36% disapproval) as of Dec 16-20. Pretty low by Putin’s usual stratospheric standards, but still figures that most Western politicians would give their right arm for; nor are they all that unprecedented (Putin’s approval ratings, according to Levada, were in the mid-60%’s for most of 2005). According to VCIOM, Putin’s approval fell from 61% on Nov 30, to just 51% in Dec 10. I don’t know why the big (and consistent) discrepancy with the Levada results – VCIOM’s typically give Putin 5%-10% points lower than Levada – but I assume it may have something to do with the wording (Putin is mentioned by name in Levada’s polls, whereas VCIOM only refers to his official position).
Likewise, both pollsters indicate that Putin remains politically hegemonic in Russia. As of Dec 10, VCIOM says 42% of Russians would vote for Putin, compared to 11% for Zyuganov, 9% for Zhirinovsky, and 5% for Mironov (everyone else has inconsequential figures); accounting for the 27% of people who won’t participate, haven’t decided, or will spoil their ballots, this gives Putin a 57% victory in the first round. As for Dec 20, Levada says that 36% will vote for Putin, compared to 7% for Zhirinovsky, 6% for Zyuganov, and 2% each for Mironov and Prokhorov. Bizarrely, it includes Medvedev – who won’t participate – in the poll; let’s (conservatively) assume that two thirds of his 3% can be added to Putin’s to give him 38%. Accounting for the 42% of people who won’t participate or haven’t decided, this gives Putin a first round victory of 65%.
Despite everything that’s happened, I can only conclude that the polls – generally speaking, very reliable predictors of Russian elections – indicate that Putin is on course to win the 2012 Presidential elections in the first round. This will doubtless raise howls of protest from the liberals, the Western media, and various hypocrite do-gooders like Hillary Clinton decrying the fairness of the elections. That is why I think that Putin’s suggestions to install web cameras at each of Russia’s 95,000 polling stations – despite its vast expense, at half a billion dollars – is in fact a very good idea so as to forestall as much as possible the inevitable and incipient attempts to delegitimize his return to the Presidency.
Obviously, this isn’t my last word on the fraud, opinion polls, and what they all mean for political developments in Russia. I hope to have another article out in a few days, as well as a blog post summarizing the various statistical analyses of the scale of fraud in the elections (as I promised to in a comment to a post at Mark Chapman’s blog).
PS. This is kind of whimsical, but I would also point out that even the informal poll at this blog – with its Gaussian shape and peak at the 2-10% fraud option – also supports the 5% fraud thesis. 😉
* I remember Julia Ioffe tweeting something disparaging about “throwing around numbers like confetti” during Putin’s Q&A, as if that’s a Bad Thing and inappropriate in an argument. I beg to differ, and consider that numbers and statistics are far more important than the (inevitable pro-liberal) rhetoric that she seems to favor.
** These figures may be off by about 1% point because I could only find the VCIOM figures rounded up to the nearest whole digit. But this is irrelevant anyway because their margin of error is 3.4%.
*** I know that some of the commentators at my blog treat opinion pollsters, especially independent ones like Levada, with suspicion because of their foreign funding. Guys, be realistic. That might be true and the director of Levada is no fan of Putin, but on the other side of the aisle there are people arguing that Russian opinion pollsters are in fact stooges of the Kremlin because their results usually suggest that the results are, in fact, fair. Let’s have less conspiracy theories and more Occam’s Razor: that they are, in fact, honest, but neither are they absolutely impeccable due to their error margins and possible sampling mistakes.