Install

Get the latest updates as we post them — right on your browser

. Last Updated: 07/27/2016

Why Polls Do Not Match Up

With most pollsters agreed one month before the June 16 presidential polls that President Boris Yeltsin has made up a huge shortfall against his Communist Party challenger and now even enjoys a narrow lead over Gennady Zyuganov, Nugzar Betaneli is proving to be a very awkward fly in the ointment.


Betaneli is the director of the Institute with 43 to 45 percent of the vote, against Yeltsin's 25 percent.


Polling is a relatively new science in Russia, and findings over the past five years have often proved startlingly wrong. In the parliamentary elections of 1993, for instance, most pollsters' predictions of an overwhelming victory by radical reformist Yegor Gaidar were confounded as Vladimir Zhirinovsky's ultranationalists swept to victory.


In last December's State Duma elections, most polls were closer to the mark, rightly predicting a victory for the Communists at the expense of the reformist democrats. But here again, they underestimated both the extent of the Communist landslide and the enduring influence of Zhirinovsky, while overestimating the strength of the pro-government Our Home Is Russia.


In both cases, however, Betaneli was one of the few pollsters to get it right. In 1993, he predicted the success of Zhirinovsky, and in 1995 his prognosis was again one of the most accurate.


Yet Betaneli is not the only pollster with a good track record. CESSI, for example, which produces The Moscow Times/CNN poll, also got it right in 1993. It is in his current figures that Betaneli is an anomaly.


Part of the reason may be that in comparing Betaneli's figures with those of the rest of the pollsters, one is comparing apples with oranges.


When the other pollsters say Yeltsin and Zyuganov are neck and neck, with support ranging from 20 to 30 percent each depending on the poll, they mean that this is the number of people who said they had made up their minds to vote for each candidate. Betaneli's May 2 figures for the number of people who had resolved to vote for Yeltsin and Zyuganov were 16 percent and 26 percent respectively.


Where Betaneli differs is that he goes on to say that, according to "social psychological prognoses," Zyuganov has the "possibility to receive 43 to 45 percent of the votes of Russian voters in the elections on June 16, 1996." The equivalent calculation for Yeltsin was only 16 to 20 percent.


Given how many voters have not yet made up their minds on whether or how to vote, Betaneli's extrapolation could well be right, but it is hard to tell because he declines to discuss his methodology.


And according to polling experts, methodology is the key to success.


"The first source of likely error is the questions themselves and the context in which they are put," said Nick Winkfield, a partner in the British MORI polling and market research organization. "The second is the sampling." Winkfield said the wording of the questions and the tone in which they were put could influence the answers.


"If you have put a whole lot of questions about economic performance, and then you ask people who they are going to vote for, you could well get a very different reply to what you would get if you had been talking about, say, law and order," he said.


It is crucial, in Winkfield's view, to choose a sample of the electorate that accurately reflects the differences between rural and urban voters, their level of education and access to the media, as well as regional voting trends. It is also useful to find out if the polling organization is carrying out any separate market research, as that would also affect its choice of population sample.


"If the organization is not tied in any way to consumer marketing, then that's fine. But otherwise, nobody is interested in setting up a research infrastructure where no commercial business exists," Winkfield said. And whether there is active commerce, people are more likely to vote for Yeltsin.


If those are the main criteria for accurate surveys, then Russia's main polling organizations seem to be getting it about right. They provide extensive breakdowns on the numbers of people interviewed, the questions asked, the timing of the interviews, the age, sex, education and income of respondents, as well as separate analysis of the regions and size of population centers.


The exception to this rule is Betaneli's institute, which confines itself to publishing its conclusions, while providing minimum data on how these were reached. Betaneli said last week's findings were the result of interviews conducted between April 27 and May 2 among 6,000 respondents from all social backgrounds in 62 regions and cities across Russia, including Moscow and St. Petersburg. But he gave no details of what questions were asked or any breakdown of the responses."We know nothing about his methodology, so it is impossible to explain the discrepancies between his findings and ours," said Yury Levada, director of the All-Russia Center for the Study of Public Opinion, or VTsIOM.


Vladimir Andreyenkov, director of the CESSI organization, which shared Betaneli's success in predicting the 1993 Zhirinovsky victory but is now completely at variance with Betaneli's view that Zyuganov is well ahead, said this week he was at a loss to explain the discrepancy.


"Betaneli does not provide opinion polls, but only his prognoses. His actual findings are not published -- and remain a secret even to us," Andreyenkov said.


Betaneli declined to give any opinion on why his conclusions differed so widely from those of his rivals. "I don't wish to make any comment on how they conduct their surveys," he said. "And as far as I am concerned, just look at my record."


All the pollsters agree on the certainty of both Zyuganov's and Yeltsin's going through to the second-round run-off. And CESSI, VTsIOM, the Institute of Public Opinion, the ROMIR polling group -- all say that the latest findings indicate a slight advantage for Yeltsin.


Andreyenkov said the final picture could change dramatically over the last weeks of intensive campaigning, while the outcome of the second round would depend very much on maintaining voter interest at the height of summer.


"Second-round turnouts are always smaller. Add to that the summer factor -- particularly for the wealthier voters abroad or at their dachas -- and Yeltsin could have quite a problem on his hands," he said.


According to Betaneli, it is far too early to give any accurate prognosis about the outcome of the second round.


"People are not psychologically prepared for the question. They are still focused on the first round. So whatever they say now about their intentions may have nothing to do with how they actually vote in the event when they see how the first round has turned out," he said.


"But according to the indications I am getting, Zyuganov is still in the lead," he said.