Translated material
Black magic of numbers
Sociologists do not understand why their forecasts coincide with the election results
The features of electoral sociology were discussed by representatives of leading sociological services at a round table in the House of Journalists. The reason for the discussion was pre-election forecasts created on the basis of sociological surveys, which in the parliamentary elections on December 2 surprisingly coincided with the official data of the Central Election Commission. The discrepancies did not exceed half a percent.The Central Election Commission, naturally, immediately seized on this data and interpreted it as evidence that the elections were fair and fair. Nevertheless, many experts have no doubt that the parliamentary elections were not without serious violations. In this case, it is not clear how to explain the accuracy of the predictions. “What did sociologists predict - the results of the Central Election Commission or the actual electoral behavior of citizens?” - asked the General Secretary of the Union of Journalists, Igor Yakovenko, who acted as the host of the round table. It was this riddle that the luminaries of sociological science tried to solve.
To begin with, we remembered that not everywhere the forecasts and results of exit-polls were in perfect agreement with the official results. “There is information about individual significant deviations of the order of 15-18%,” said Igor Zadorin, general director of the Zircon research group. “They are unevenly distributed throughout the country, most of them are in the Volga and North Caucasus regions.” We are talking about Tatarstan, Bashkiria, Mordovia, Ingushetia, Chechnya and other republics, which on December 2, 2007 pleased the Central Election Commission with an almost universal turnout and almost absolute voting for “United Russia”.
An example of such unanimity was given by the head of the Mercator group, Dmitry Oreshkin: “In Dagestan, there is an election commission, 15 polling stations are assigned to it, 17,779 ballots were issued for them - exactly the same number of people voted, all for United Russia.” How can this be predicted? The head of the political research department of the All-Russian Center for the Study of Public Opinion, Stepan Lvov, suggested that it’s all about a special “type of political culture.” Mr. Oreshkin objected that the problem lies in the “overwhelm of administrative resources”: “Yeltsin relied on cities, and Putin and United Russia, judging by official data, relied on small towns and villages, provinces dependent on administrative resources.” . “There are two different Russias,” the analyst concluded, “one votes as best it can, the other votes as the authorities say.”
Administrative resources and specific political culture were linked together by the President of the INDEM Foundation, Georgy Satarov, who explained that in a number of national republics the “patriarch,” the elder, tells his fellow tribesmen how to vote. People check the boxes themselves, of their own free will, although they are commanded through the elders by representatives of the federal center. Deputy Director of the Levada Center Alexei Grazhdankin added that in such cases, opinion polls also show the will of voters with similar distortions, since sociologists talk with respondents after the “patriarchs”. As a result, surveys show in advance the level of abuse of administrative resources.
Mr. Grazhdankin also drew the attention of his colleagues to the fact that not all Russians are eager to share their political sympathies with sociologists. And among those who agree to participate in surveys, there are more United Russia supporters than among the population as a whole. This gives the party in power the extra percentage in forecasts, which the notorious administrative resource provides it with during voting.
But the most important reason why sociologists and the Central Electoral Commission agree on the figures is that a survey on political preferences is not yet considered an electoral forecast. To predict the election results, the resulting ratings are also multiplied by special coefficients. Each sociological service has its own, but they have one thing in common: they are strictly tied to the official results of previous votes. And no matter what is included in the data of the Central Election Commission - administrative resources, falsification, stuffing, sociologists take into account these additional circumstances through coefficients. But the possibilities of falsification are not limitless. According to Dmitry Oreshkin’s calculations, regions “with a special electoral culture” brought about 4 million “extra” votes to the party in power, that is, approximately 3-5%. “In 1996, I tried to estimate the stuffing - it also turned out to be 4 million,” noted Igor Zadorin. “This is a kind of sacramental figure that limits the possibility of amendments.”
In conclusion, Georgy Satarov and Igor Yakovenko proposed making the mechanism for compiling electoral forecasts more transparent, so that it would become clear where sociologists get their numbers, and there would be no reason to suspect them of collusion with the Central Election Commission.