PICK YOUR POLL: PUTIN'S PARTY WINS BIG

C O N F I D E N T I A L SECTION 01 OF 03 MOSCOW 005540 SIPDIS SIPDIS E.O. 12958: DECL: 11/26/2017 TAGS: PGOV PHUM SOCI RS SUBJECT: PICK YOUR POLL: PUTIN'S PARTY WINS BIG Classified By: Political M/C Alice G. Wells. Reason: 1.4 (d) -------------- Summary -------------- ¶1. (SBU) According to the final public opinion polls published one week before the December 2 Russian Duma elections, United Russia (YR) will garner the vast majority of votes and the Communist Party (KPRF) will easily overcome the seven percent threshold to entry into the Duma. The predictions of the three big polling agencies differ on whether A Just Russia (SR) and The Liberal Democratic Part of Russia (LDPR) will enter the Duma. The data indicate no chance for the opposition Yabloko party and the Union of Right Forces (SPS) to garner enough votes for a place at the table. Despite questions raised about polling bias in Russia, the three agencies--using different sampling and questions--have captured the same trends as well as the same bottomline: United Russia's overwhelming win. End Summary. -------------- Polling Sources & Bias -------------- ¶2. (SBU) Three established and competent firms conduct routine political and sociological polling within Russia: the All Russia Institute for Public Opinion (VTsIOM),the Levada Center, and the Foundation for Public Opinion (FOM). Other research organizations in Russia could also conduct electoral polling; however, these organizations do not routinely make their results public. VTsIOM was established by Yuriy Levada in 1987 under Gorbachev during perestroika. In 2003, the Russian government took over VTsIOM and installed a Kremlin approved director and governing board. Yuriy Levada, who established the center, left at that time. Most of the researchers followed him to his new public opinion organization now known as the Levada Center. Since that time VTsIOM has been a wholly-owned government enterprise. Andrey Mukhin of the Center for Political Technology contends that government ownership has biased VTsIOM in favor of the Kremlin. ¶3. (U) In its November 5 issue, The New Times published an investigation into VTsIOM. While the thrust of the article was the company's use of off-shore accounts in Cyprus and the British Virgin Islands, it also accused the organization of using leading questions to obtain Kremlin-specified results. For example, the article pointed out that when asking about the Union of Right Forces (SPS),VTsIOM interviewers asked respondents if the party was a party of reform and democracy or a party of rich, antipatriotic oligarchs. Similarly, when asking about SR, VTsIOM asked the respondents' opinion about the inclusion of millionaires and criminals on the SR party list. ¶4. (C) Some analysts, including Mukhin, maintained that the data from all three main polling and public opinion organizations must be taken with a grain of salt and perhaps a suspension of disbelief. Mukhin told us that each organization, to one degree or another, is controlled by the Kremlin. Although the data are reliable -- they will produce the same results time after time -- they lack validity (the ability to produce results that represent reality). VTsIOM, Levada Center, and FOM are under pressure to produce numbers that conform to Kremlin interests. In a November 15 conversation, Mukhin stated that the polling numbers for United Russia are basically accurate; however, the numbers for other parties do not represent how these other parties will perform during the election. Mukhin reported that valid and reliable polling for the Duma election exists; however, the polls are done by private firms for private individuals or organizations. The results are not for public consumption. In other words, he felt that neither the media nor the public can see how well parties other than YR are doing in the election campaign. ¶5. (C) In previous conversations Mukhin explained that the Kremlin has a keen interest in how the public feels. In order to get accurate information, the Kremlin has its own polls that provide a more realistic assessment of public sentiment but are not released to the public. Mukhin believed that the results of these polls to a large extent prompted Putin's surprise decision to head United Russia's party list. ¶6. (C) Leonid Sedov of the Levada Center provided a different assessment of the survey organizations in a November 16 conversation. Sedov worked originally for VTsIOM but also left to help set up the new Levada Center. He has been involved in public opinion research for more than 20 years. He contested any notion that the Kremlin has MOSCOW 00005540 002 OF 003 influence over the results of their surveys. -------------- A Two-Party Duma? -------------- ¶7. (U) The numbers produced by each organization follow the same trends and produce roughly the same results. At a press conference on November 26, VTsIOM released its latest polling data and predictions of results showing United Russia (YR) receiving 62.1 percent of the vote. On November 23, Levada Center released similar data showing 67 percent of votes for YR. On November 22, FOM released data indicating 53 percent of voters would vote for YR. These numbers of course represent only those who intend to vote. Each organization conducted the surveys November 17-18. ¶8. (U) The three organizations have reported numbers that have been internally consistent from week to week although they have seldom matched. The trends that each organization have reported have been roughly consistent. Following the October 1 announcement that Putin would head the United Russia party list, YR saw an increase in the polls from each organization ranging from six to ten percent. Beyond this single radical change, the polling numbers for each party have not changed significantly since the start of the election campaign. ¶9. (U) Each organization reported that KPRF would receive the second largest vote come December 2. KPRF would easily overcome the seven percent barrier to Duma representation; even if it did not, as runner-up it would enter the Duma because the Russian Constitution guarantees a minimum two parties in the Duma. According to VTsIOM, the party would receive 12.2% of the vote. The Levada Center reported 14 percent for the Communists while FOM estimated their results at 10 percent. ¶10. (U) Just Russia (SR) and LDPR are facing an increasingly tough road ahead in the election campaign. SR, which at one time polled as high as 15 percent, now faces dismal ratings in the single digits. FOM reported that SR would receive from seven to eight percent of the vote, Levada Center reported four percent, and VTsIOM reported seven percent. LDPR polled roughly equal to SR in many of the previous polls. However, the latest polls have shown LDPR in a better situation. VTsIOM's most recent predictions indicated LDPR receiving eight percent of the vote while Levada Center estimated six percent. FOM predicted that LDPR would poll as well as KPRF or better with 11 percent. ¶11. (U) The story for all other parties appeared grim. FOM found that the smaller parties barely registered in the polls. VTsIOM and Levada Center noted one or two parties that rated above one percent, but generally parties like the Union of Right Forces (SPS),Yabloko and the Agrarian Party polled at around one to two percent (slightly higher than the previous week when all polled at less than one percent). Given that each poll had a statistical error of plus or minus three percent, the three organizations, for the December 2 elections at least, unanimously pronounced these parties dead on arrival. -------------- Those Darn Statistics -------------- ¶12. (C) According to Mukhin and Sedov, Russian polling agencies use rather complicated strategies which are hinted at but never fully discussed in the results. Each organization has divided the Russian Federation into blocks which may or may not correspond to political boundaries. The blocks also may or may not be the same from one polling organization to another. The organizations in turn sampled from these blocks but not with the same probability. Certain types of blocks such as small villages might have been oversampled to adjust for regional variations. The researchers in each firm sampled respondents from within these blocks. At each stage, each possible respondent had a different probability of being sampled. These differing probabilities must be taken into account when reporting the results; otherwise the reported results would be biased. ¶13. (U) Based on the results from FOM, VTsIOM, and Levada Center, each organization has conducted their polling with some consistent, built-in bias. In other words, each poll has produced roughly the same results from one round to the next with each organization's numbers differing from the others by approximately the same amount. The trends that each polling agencies detected was also detected by the other two. In other words, the differences in data collection methods have produced results that differ in somewhat MOSCOW 00005540 003 OF 003 predictable fashion, but that provide overall reliable results. ¶14. (SBU) Mark Twain is said to have popularized the quote from Disraeli saying there are only three kinds of lies: "Lies, damn lies, and statistics." Each polling agency provides a rough estimate of statistical error to account for the problems inherent in any sampling of a population. However, this estimate is a mathematical estimation that does not account for the non-mathematical biases in the data collection methods. Saying that United Russia will receive 62.1 percent of the vote sounds meticulous and scientific, but the truth is that United Russia will receive anywhere between 59 percent (very close to some of the lowest estimates of its performance) and 65 percent if one uses the standardized error estimates provided. Statistical interpretations can lie or even just mislead. They lie by providing a false sense of certainty and exactness, by not taking into account important procedural factors that can alter the results, and by analyzing trends and differences that are statistically undetectable. ¶15. (SBU) The value of the Russian polling data, then, is in its ability to provide broad statements about political opinions and voting behaviors. Following Putin's October 1 announcement that he would head YR's national party list, YR received a significant increase in support while SR suddenly found itself unexpectedly struggling. From the big three polling organizations, YR clearly will get a simple, if not constitutional, majority (one where it can change the constitution without the support of another party). Also, KPRF will be the de facto opposition party in the next Duma. Both SR and LDPR are still fighting to get over the seven percent hurdle, but it appears impossible for Yabloko or SPS to reach that threshold. -------------- Comment -------------- ¶16. (SBU) The vastness of Russia, its economic diversity, and its unique population dispersion create particular challenges for polling here. Current polling numbers indicate a potential shift in Russian politics come December as YR gains an even tighter grip on power, the KPRF becomes the only viable opposition and the other parties die off in a mass extinction. What is noteworthy is that the three main organizations have reported the same general results although they differed in the details. While the Kremlin might have interfered in VTsIOM's polls results, or all three polling organizations may have deliberately used leading questions to bias respondents' answers, given the context of three different surveys reporting the same results, it is difficult to determine if these techniques actually changed the respondents' answers. The polls could not miss the exceptional support currently enjoyed by United Russia. Similarly, the polls could not dismiss KPRF as the second party. If Kremlin interference in the polls was designed to mislead the public and opinion makers, then it could only be done in the margins. If the apparent faith in United Russia were not real, such a large error could not stay hidden. BURNS