In the News · November 1, 2024
Edison Research 2024 Republic of Georgia Exit Poll
By Edison Research
Analysis of discrepancy between exit poll and CEC results
November 1, 2024
Edison Research has completed a review of its exit polls conducted on the day of the recent parliamentary elections in Georgia. This review was undertaken following allegations that voter fraud was a key determinant in the election outcome.
Edison has been conducting national exit polls in Georgia since 2012. All of Edison’s previous exit polls accurately supported the official vote showing Georgian Dream as the winning party. The average understatement of the Georgian Dream vote over five exit polls before 2024 standing at under 2%.
However, the official outcome of the 2024 parliamentary election put the ruling Georgia Dream party’s popular vote at 54% – or 13 percentage points greater than Edison’s exit poll estimate on the evening of October 26.
Edison Executive Vice President Rob Farbman commented, “The 13-point difference between Edison’s estimate and the official result of 54% for Georgian Dream cannot be explained by normal variation alone and suggests local-level manipulation of the vote.
“Edison’s review found that the deviation from statistically expected results was widespread but most pronounced at specific polling locations in rural areas. These locations are likely to have had the most significant vote manipulation at the polling location level.”
Edison’s exit poll models use a combination of data sources. Edison’s estimates use Election Day vote tallies from exit poll questionnaires, past vote history for each voting location, and weighting adjustments based on differing response rates by demographic groups and past vote. Edison’s Election Day process also includes quality checks of the data by polling location to remove or reweight any data that appears to have issues or errors that will make estimates less accurate – these include interference by an election official, interviewer error, bad weather, or evidence of fraud.
In its exit poll on October 26, Edison Research implemented quality control measures that identified polling locations with significant deviations from statistical norms when compared to historical voting patterns.
Edison also conducted 12 nationally representative surveys in Georgia in 2023-24. These polls suggested that Georgian Dream would under-perform their 2020 vote by around 10 points. Instead, Georgian Dream exceeded their 2020 numbers by 6 points.
The exit poll models used in Georgia in 2024 are the same models used in previous Georgian elections. They are also the same models Edison Research will use in the upcoming U.S. election exit polls for ABC, CBS, CNN, and NBC.
For further information contact Rob Farbman at rfarbman@edisonresearch.com
Get our latest insights delivered to your inbox.
Analysis of discrepancy between exit poll and CEC results
November 1, 2024
Edison Research has completed a review of its exit polls conducted on the day of the recent parliamentary elections in Georgia. This review was undertaken following allegations that voter fraud was a key determinant in the election outcome.
Edison has been conducting national exit polls in Georgia since 2012. All of Edison’s previous exit polls accurately supported the official vote showing Georgian Dream as the winning party. The average understatement of the Georgian Dream vote over five exit polls before 2024 standing at under 2%.
However, the official outcome of the 2024 parliamentary election put the ruling Georgia Dream party’s popular vote at 54% – or 13 percentage points greater than Edison’s exit poll estimate on the evening of October 26.
Edison Executive Vice President Rob Farbman commented, “The 13-point difference between Edison’s estimate and the official result of 54% for Georgian Dream cannot be explained by normal variation alone and suggests local-level manipulation of the vote.
“Edison’s review found that the deviation from statistically expected results was widespread but most pronounced at specific polling locations in rural areas. These locations are likely to have had the most significant vote manipulation at the polling location level.”
Edison’s exit poll models use a combination of data sources. Edison’s estimates use Election Day vote tallies from exit poll questionnaires, past vote history for each voting location, and weighting adjustments based on differing response rates by demographic groups and past vote. Edison’s Election Day process also includes quality checks of the data by polling location to remove or reweight any data that appears to have issues or errors that will make estimates less accurate – these include interference by an election official, interviewer error, bad weather, or evidence of fraud.
In its exit poll on October 26, Edison Research implemented quality control measures that identified polling locations with significant deviations from statistical norms when compared to historical voting patterns.
Edison also conducted 12 nationally representative surveys in Georgia in 2023-24. These polls suggested that Georgian Dream would under-perform their 2020 vote by around 10 points. Instead, Georgian Dream exceeded their 2020 numbers by 6 points.
The exit poll models used in Georgia in 2024 are the same models used in previous Georgian elections. They are also the same models Edison Research will use in the upcoming U.S. election exit polls for ABC, CBS, CNN, and NBC.
For further information contact Rob Farbman at rfarbman@edisonresearch.com