[Politics, That Day...] Everyone Holds Their Breath in Front of the TV at '6 PM', the Secret Box of the General Election Opens

April 15 21st General Election Broadcast 3 Companies Exit Poll Results Draw Attention... Why the General Election Became the Graveyard of Polling Agencies

[Asia Economy Reporter Ryu Jeong-min] ‘Politics, On That Day…’ is a series planning corner that looks back on Korean politics through the ‘recollection of memories’ related to notable scenes, events, and figures.

▲KBS, MBC, SBS Broadcast 3 Companies Exit Poll

“I actually knew the exit poll results.” This was said by a senior opposition politician who left a ‘famous video’ during the 20th general election on April 13, 2016. Although he appeared to smile before the 6 PM exit poll announcement and then his facial expression seemed to stiffen afterward, he explained that this was just how it appeared on the broadcast screen.

The most tense moment in the general election, which comes every four years, is 6 PM on election day. This year as well, on April 15 at 6 PM, many people are expected to hold their breath in front of their TVs waiting for the exit poll results from the three major broadcasters.

Elections are a battlefield of intense information warfare. The exit poll results, which determine the victory or defeat, are everyone's concern. The three major broadcasters keep the results strictly confidential until the announcement, but sometimes the party information lines’ radar nets catch wind of them.

However, even if they know the results, they cannot leak or announce them externally. The three broadcasters spend a large amount of money conducting nationwide opinion polls in 253 constituencies, and if the results are announced before the official announcement by the media company, legal penalties may be imposed.

On the 29th, about two weeks before the 21st National Assembly election, promotional materials encouraging voting were installed at Cheonggyecheon in Seoul. Photo by Kim Hyun-min kimhyun81@

This year, the nationwide expected seats of major parties announced by the three broadcasters can be publicly disclosed by other media at 6:10 PM, and the expected vote rates for specific constituencies like Jongno in Seoul can be announced at 6:30 PM. Although the three broadcasters release a flood of information related to the election situation starting at 6 PM, other media cannot immediately receive and use this information.

The reason why the three broadcasters’ exit polls attract attention is the belief that the victory or defeat can be confirmed immediately after the polls close. Especially in presidential elections, the accuracy of predicting the winner is high. But general elections are different.

The three broadcasters’ exit polls do not predict exact numbers, such as Party A winning 140 seats and Party B winning 120 seats out of 300 total seats. Instead, they predict within a certain range, such as between 125 and 140 seats or between 105 and 125 seats.

The problem is that even with a wide range set, the actual seats may differ. The 20th general election held just before was such a case. The three broadcasters’ general election predictions are not decided solely by exit polls. They comprehensively analyze various opinion polls and public sentiment trends released before the exit polls to produce the results.

Even when conducting joint exit polls, the expected seats differ by broadcaster for this reason.

On the 13th, about a month before the 21st National Assembly election, a voting participation promotional poster installed by the Seoul Metropolitan Election Commission is displayed on the screen door of Chungmuro Station on Subway Line 3 in Jung-gu, Seoul. Photo by Jinhyung Kang aymsdream@

In the 20th general election, KBS predicted Saenuri Party 121-143 seats, Democratic Party 101-123 seats, and People’s Party 34-41 seats. MBC predicted Saenuri Party 118-136 seats, Democratic Party 107-128 seats, and People’s Party 32-42 seats. SBS predicted Saenuri Party 123-147 seats, Democratic Party 97-120 seats, and People’s Party 31-43 seats.

The three broadcasters’ predictions implied that the Saenuri Party would find it difficult to achieve a majority but was likely to become the largest party in the National Assembly, the Democratic Party would be the second-largest party with around 110 seats, and the People’s Party would perform well enough to form a parliamentary negotiation group (with 20 or more members).

The People’s Party part of the three broadcasters’ predictions was generally accurate, but the Saenuri Party and Democratic Party parts differed from the actual results. The final count was 123 seats for the Democratic Party and 122 seats for the Saenuri Party.

Although the three broadcasters set a range of about 20 seats, the Saenuri Party’s seats were at the lowest end of the prediction or even outside the range. Meanwhile, the Democratic Party’s predicted seats were at the high end or outside the range. Contrary to expectations, the Democratic Party became the largest party in the National Assembly.

Overall, the Democratic Party’s seats were underestimated, and the Saenuri Party’s seats were overestimated. The reason the three broadcasters had to present such results was that various opinion polls released before the general election showed such outcomes.

On the 13th, when the 7th nationwide local elections and the by-elections for the National Assembly were held, election staff at the Yeongdeungpo District Election Commission counting center, set up at the Yeongdeungpo Multipurpose Badminton Gymnasium in Seoul, were moving ballot boxes. Photo by Hyunmin Kim kimhyun81@

The reason the general election is considered the graveyard of opinion polls is that accurately predicting the situation in all 253 constituencies nationwide is not as easy as it sounds. Considering a sample size of 500 per constituency and the margin of error, it is not surprising if close races yield results different from predictions.

If the so-called ‘hidden votes’ are added, the predictions can be completely off. What about the 21st general election? If we only look at the 6 PM exit polls, can we at least know who will be the largest party in the National Assembly? Considering the advancement of polling techniques and various ‘correction devices’ by broadcasters, accuracy is expected to improve, but whether the hit rate will be high this time is questionable.

It may be reasonable to treat results for specific constituencies as reference material. If a candidate predicted to be a likely winner in a broadcaster’s forecast is overturned in the actual count, the shock for the candidate and supporters can be significant.

Nevertheless, the reason why 6 PM on election day is eagerly awaited is that it scratches the most itching spot. As long as the belief that it will resolve the curiosity about the general election’s outcome remains, 6 PM on election day will inevitably be the hottest time.