An “exit-poll” is a new notion in Ukraine. For a long time, it has been exotic even for journalists (who insisted on calling it “exit-pool”), let alone people in the street. Yet the latest developments have made the general public politically conscious, and masses of demonstrators have adopted it as part of their lexicon.

Exit-polls have been conducted in Ukraine since 1998, and their results have never diverged considerably from the Central Election Commission’s official data, which added legitimacy and cogency to the election outcomes.

It was in the mayoral elections in Mukachevo that the exit-poll data and the official results differed dramatically - by 60%, given the cumulative difference between two candidates. Gearing up for the use of the “Mukachevo scenario” in the nation-wide elections, the authorities had to decide what to do about the exit-poll, due to take place as usual. Events in Georgia, where the exit-poll laid bare ballot rigging in favour of the “party in power” and triggered the “rose revolution”, even further aggravated the problem facing the authorities.

The mass media launched a campaign against exit-polls long ago. At first, though, the “single candidate’s” team did not know “how to manipulate” (as that candidate told the press in an interview) this technique. Soon they found a solution: to conduct an alternative exit-poll or, rather, a series of them. It was essential for them that by the end of voting day, October 31, all the exit-polls should yield homogeneous results, presenting Viktor Yanukovich as the winner. That is why at 6 pm on that day the Internet sites maintained by the Social Democrats and the Donetsk group, reported that all four exit-polls were showing Yanukovych’s lead.

The attempt failed. The truth is that a professional survey (with correctly designed methodology, accurate sampling and a clear “margin”) is resistant to falsification; it is carried out independently of its organizers’ preferences and produces results that cannot be pre-arranged.

The Russia-sponsored Public Opinion Fund gave up its survey, having pronounced it abortive because of “a high rejection rate” (which, nonetheless, did not prevent Gleb Pavlovsky from making public its results favouring Yanukovych). Members of the National Exit-poll Consortium used two different methodologies - those of personal interviews (Socis and Social Monitoring) and of anonymous questionnaires (Kyiv International Institute of Sociology and the Razumkov Centre). Their outcomes varied: the anonymous questionnaires showed Viktor Yushchenko’s advantage over Viktor Yanukovych (44.6% versus 37.8%); according to the open interview, Yushchenko also was 1.8 points ahead of Yanukovych. The only exit-poll whose results disagreed with the above was that conducted by a group of universities headed by Olexander Yaremenko, which gave a noticeable prevalence (of about 4%) to Viktor Yanukovych.

Unfortunately, the night after the elections proved tragic for national sociology, as our former Consortium fellows (Socis and Social Monitoring) never published the results they got but replaced them with other numbers, surprisingly similar to those of Yaremenko’s exit-poll and the foiled but announced Public Opinion Fund’s exit-poll.

The events that followed, the rip in the Consortium ensuing [Socis Director] Mykola Churilov’s mysterious “weighting” of the poll results that turned them upside down, were described in detail in numerous media outlets, including ZN (see Tetiana Silina’s article “Exit-poll: the torments of Tantalus”).

I would like to state again, emphatically, that in the first round of the elections, three nation-wide exit-polls conducted independently confirmed Viktor Yushchenko’s advantage.

More than that, the interviews conducted by the four Consortium members simultaneously with the national exit-poll included a question on the respondents’ likely choice in the run-off. All four firms received answers indicative of Viktor Yushchenko’s likely victory. All four firms polled their respondents by the customary method of open interview; none used ballot-boxes this time. The outcomes were very close, which enabled the sociologists to forecast Viktor Yushchenko’s lead in the run-off.

By the time of the run-off, the Consortium lost half of its members, having retained the Razumkov Centre and Kyiv International Institute of Sociology, whose efforts were coordinated by the Democratic Initiatives Fund. Socis and Social Monitoring claimed they would carry out their own exit-poll, resonantly entitled “People’s Choice”.

The Consortium members prepared for their run-off exit-poll even more meticulously than for the first one. We invited international experts from Russia and Poland who have themselves conducted dozens of exit-polls. We improved our sampling methodology and anonymous surveying technique (used in various countries) whereby respondents fill in questionnaires and drop them into a box. We proclaimed our absolute transparency, and the openness of both our methodology and funding to the sociological community and general public.

By and large, the national exit-poll in the run-off was productive, with a rejection rate amounting to 27%, which is an acceptable rate by all international standards. The run-off exit-poll results obtained by the Razumkov Centre and Kyiv International Institute of Sociology were very close: they coincided by decimal percentage points for Yanukovych and deviated by 1.5% for Yushchenko, which testifies to the credibility of the data produced.

According to the press release published on 21 November, about 30 thousand respondents were surveyed at 460 polling stations. 54% of the respondents voted for Viktor Yushchenko, 43% for Viktor Yanukovych and 3% voted against both. The statistical error does not exceed 2%, accessibility level being 73%.

It is true, the results did not reflect the CEC data on different turn-out in some oblasts (respective data were not available at the time). The turn-out in eastern oblasts, according to the CEC, was extraordinarily high, but it is not within sociologists’ purview to ascertain the truthfulness of the CEC data. The exit-poll methodology requires that the results obtained should be adjusted with due regard of the CEC turn-out data for various regions of Ukraine (raising the sample share of the oblasts with higher turn-out).

Therefore, our sociologists weighed the above results to incorporate the CEC data. Yet even with 96.65% turn-out in Donetsk Oblast and 89.53% turn-out in Luhansk Oblast, the adjusted outcome is still very different from the results that CEC announced on 24 November. Even this evidently misrepresented turn-out data for the Eastern oblasts did not have a drastic effect on the situation - the gap between the candidates narrowed to 8%.

The remarkable discrepancy between the CEC and exit-poll data can be accounted for not only by the artificially boosted turn-out in Eastern oblasts (which exit-polls can hardly capture), but also by ballot-stuffing in favor of the governmental candidate in other regions where the voting was not homogeneous. It is understandable, given that the sociologists interviewed voters at the exit of polling stations and received different responses, while the already-ticked ballots were all in favour of one candidate.

The regions with heterogeneous voting, i.e. oblasts in Central Ukraine, showed the greatest divergence. Although our survey is not representative for individual oblasts, special attention should be paid to the oblasts where the difference between the CEC and exit-poll data was the most pronounced: Trans-Carpathia, Kirovohrad, Sumy, Kherson, and Khmelnystky. One can conclude, therefore, that ballot-rigging was not geographically confined to the Eastern part of Ukraine, but was a common and systematic practice.

What happened to the other exit-poll projects? Did they contrive to substantiate the “single candidate’s” victory?

Olexander Yaremenko, “Ukrainian Exit-poll” Project Manager, followed suit of the Russia-based Public Opinion Fund, and pronounced his poll failed due to… “a high rejection rate”.

Social Monitoring did carry out its poll, and Olga Balakireva, Firm Manager, announced Yushchenko’s victory with 49.7% versus Yanukovych’s 46.7%. Yushchenko’s more modest advantage over his competitor, as compared to the National exit-poll data, is explained by the chosen methodology: an “open interview” format suggests that the interviewer knows the respondent’s answer. Under the strained political conditions of today, some voters could feel reluctant to disclose their choice (particularly, if it was in the opposition candidate’s favour).

As for the Socis Centre headed by Mykola Churilov, its exit-poll was yet another detective story. At first, the Centre made public the data, according to which Yushchenko got 49.1% and Yanukovych got 45%. It must have been inadmissible for the Centre’s clients. So, upon weighting those data with a special “Churilov factor”, those data were reversed: Yushchenko - 46.7% and Yanukovych - 48.2%. I was amazed to see Mr Churilov later denounce these “weighted” data and explain at length that his exit-poll was unsuccessful and should be disregarded. His interviewers, purportedly, quit at 5.30 pm (it was too dark) and the district election commission members forced them out of polling stations, and the rejection rate shot up to 30%, etc, etc, etc. Finally, Churilov claimed that the sampling error went beyond the sociologists’ control and no conclusions on the citizen’s choice could be drawn based on the obtained data. However, the clients would not let go of him. As MP Nestor Shufrych said, “the exit-poll by this sociologist, what’s-his-name, proves that Yanukovych won”. What else could he have said? They needed at least one “exotic poll” as a gift to Mr Putin, for him to greet the “correct” winner!

Where are we now that the second round of a vibrant exit-poll process is over? According to three exit-polls (Razumkov Centre, Kyiv International Institute of Sociology and Social Monitoring) Viktor Yushchenko won in the run-off. Of course, exit-polls cannot be used as a legal argument, but their results provide ample grounds for a careful investigation.

I have a dream: to live in a country where exit-polls serve as a research tool for gauging citizens’ electoral preferences and where sociologists do not have to summon all their courage and civil awareness to publicize their results. I hope we will live in such a country when we next go to vote. I hope the CEC and the exit-poll data will be in harmony and I will take pride in being a sociologist.