KEP: Local Election Exit Poll Data Error Attributable to Polling Agency’s Professional Negligence

Errors in Gender- and Age-Specific Voter Analysis Detected in Four Regions Including Seoul
Omission of Early Voter Predictive Data by Korea Research

The Korea Election Polling (KEP) Committee, under the Korea Broadcasting Association, has issued an official apology and pledged to prevent recurrence regarding the data error that occurred during the coverage of exit poll results for the June 3 local elections.

Korean Broadcasting Association

On June 11, KEP announced that it had detected errors in the gender- and age-specific voter tendency analysis data for certain regions during the election broadcasts of the three terrestrial broadcasters. KEP promptly launched an internal investigation and identified the cause.

The exit poll for the 9th nationwide local elections, conducted on June 3, was carried out by three polling agencies—Korea Research, KOREA RESEARCH, and Ipsos Korea—each assigned to different portions of the 16 provinces and metropolitan cities across the country. In order to ensure accurate predictions, it is necessary to combine the exit poll data collected on election day with the predictive phone survey data for early voters.

However, according to KEP’s own investigation, in the analysis of gender- and age-specific voters in the four regions (Seoul, Daegu, Ulsan, and Chungbuk) overseen by Korea Research, predictive data for early voters was omitted, resulting in reporting errors. KEP stated, “Although the final predictions for the winning candidates were properly generated by combining both datasets, in the gender- and age-specific voter analysis for each region, predictive data for early voters was omitted from the aggregation due to a clear professional oversight by Korea Research. As a result, this has caused misunderstandings among viewers in gauging public sentiment.”

KEP further emphasized, “The regions managed by KOREA RESEARCH and Ipsos Korea included the predictive data for early voters, and this issue occurred solely in the areas assigned to Korea Research. We clarify that KEP did not alter the data with any particular intent, but that the incident was caused by professional negligence at the polling agency.”

Additionally, KEP explained, “The three broadcasters had no means of verifying at the data receipt stage—immediately prior to the election broadcast—whether the delivered data had been produced according to the planned design. Nevertheless, we fully recognize our responsibility for oversight and supervision in this matter, and we will take all possible and stringent measures, including a complete overhaul of the verification system and legal action for any contractual violations, to ensure that similar incidents do not occur in the future.”

This content was produced with the assistance of AI translation services.