Declaration National Exit Poll 2010 Consortium

29.01.2010

The National Exit Poll 2010 Consortium conducted a meticulous analysis of the results of the exit poll we conducted during the first round of Presidential Elections in Ukraine on January 17, 2010.

As the organizers of this poll, we declare that, in accordance with our own findings and those of international auditors (Levada Center, Moscow, Russia and the Institute of Public Affairs, Warsaw, Poland), the National Exit Poll 2010 was conducted in full compliance with international professional and ethical standards.

Concerning the methodological problems that arose, we declare the following:

1.    Due to lack financing we were forced to reduce the number of polling locations from 300 to 240, resulting in a greater scale of error than we initially expected.

2.    The results of our poll and the results of the Central Election Commission (CEC) differed by 3.5% for Victor Yanukovych and 2.1% for Yulia Tymoshenko. Based on our analysis, we can explain 1% of the difference for Victor Yanukovych as the result of random error.

3.    The remaining 2.5% difference for Victor Yanukovych can be attributed to errors in the selection of respondents at the polling stations and to the peculiarities of voting on January 17. It may be the case that the difference in our results and official voting results was caused by our sample methodology, which did not account for the number of votes that were cast at home (the number of votes cast at home, according to media reports, was approximately 3.15%). Moreover, our field work was essentially completed by 18:00 on Election Day, while some polling stations saw voter mobilization rise right before the close of polling stations – something not recorded in previous elections.

4.    The differences in the results of the National Exit Poll 2010 and the other five exit polls can be primarily attributed to the deliberate use of a methodology that guarantees the independence of exit poll results from CEC data on voter participation and the number of registered voters. The results of the other exit polls were dependent, in one way or another, on this data. Our methodology was developed in 2004 when data provided by the CEC was questionable. It is based on international experience, adapted by Ukrainian and international experts and proved to be trustworthy during the parliamentary elections in 2006 and 2007. During our analysis, when we factored in corrections based on CEC voter turnout data, the results of the National Exit Poll 2010 reflected CEC official results.

  1. We do not have enough information about the exit polls conducted by other firms. Thus we call on them to reveal their polling methodologies and, after the elections are over, submit all the raw data to Ukraines sociological archives, which are open to professional scrutiny. This data contains unique information that is important for understanding electoral behavior and for planning the future development of the state.
  2. We would also like to inform the public that in the second round of presidential elections, all exit polls can find themselves in a situation similar to ours: the results could be vastly divergent because all the points gained within the margin of error by one candidate will be lost by the other candidate. Thus a margin of error of 2.5% can result in a 5% difference in the candidates’ results. If one candidate beats the other by only a small margin, the exit polls might show the other’s victory. 
  3. Finally, in conducting the exit poll for the second round of Ukraine’s presidential elections, we will eliminate the shortcomings of the methodology we employ and will separately show the results of the National Exit Poll 2010 factoring in CEC voter turnout data.

Ilko Kucheriv,

Director, Democratic Initiatives Foundation

Volodymyr Paniotto,

General Director, Kyiv International Institute of Sociology

Anatoliy Rachok,

General Director, Oleksandr Razumkov Ukrainian Center of Economic and Political Research

January 27, 2010