Democrat gubernatorial candidate MR Sukhumbhand Paribatra’s surprise victory in yesterday’s election has prompted debate among Thai netizens about the credibility of pollsters, whose exit polls forecast the exact opposite of the election’s actual results.
After the polls closed at 3pm yesterday, poll centers released their predictions about who would likely take the helm of the Bangkok Metropolitan Administration (BMA). All three exit polls forecast Pheu Thai candidate Pongsapat Pongcharoen as the winner.
An exit poll put together by Bansomdej Chaopraya Rajabhat University, TV Pool and Nation Channel’s Nation University, predicted that Pongsapat Pongcharoen would win with 40.02% of votes, surpassing Sukhumbhand’s 38.54%. The pollster claimed to have sent 500 volunteers to collect opinions from 20,000 voters between 8am and 1:30pm throughout all 50 districts of Bangkok.
Bangkok University’s poll predicted that Pongsapat would lead with 44.14% over Sukhumbhand’s 41.07%.
Suan Sunandha Rajabhat University’s poll predicted that Pongsapat would end the day with 43.88% of the vote to Suhkumbhand’s 39.57%.
The unofficial result from the city clerk’s office was that Sukhumbhand garnered 1,256,349 votes, exceeding Pongsapat’s 1,077,899. The inaccuracy of the exit polls irritated some Thai netizens, some of whom even called on pollsters to foreswear their professions.
“As long as pollsters did not create polls with genuineness, but set questions and fake poll results, the society should not accept them anymore,” said Twitter user @KobkarnL.
“Many pollsters seamlessly humiliated themselves as well,” said @NightPhoomin.
“People are increasingly pressing pollsters to take responsibility for their inaccuracy. For me, please stop creating such misleading polls,” said @raydo_thairath.
Moreover, prior to the election, entry polls by ABAC University and Suan Dusit Rajabhat University also picked out Pongsapat as the likely winner. ABAC poll said that Pongsapat would win over Sukhumbhand with a ratio of 45.9% to 34.1% while Suan Dusit poll registered predicted an election result of 49.01% to 39.65%. The only poll center that predicted a Sukhumbhand win was NIDA poll, leaving the ratio of correct to incorrect pollsters at 5:1.
That weekly opinion polls have been forecasting a Pongsapat victory for months does not help matters. Making the issue worse yet are memories of the general election of 3 July 2011, during which many pollsters proved themselves untrustworthy by leaving momentous errors in their predictions of seats won by both the Pheu Thai and Democrat Parties. Among the most notorious was a Suan Dusit exit poll that indicated that the Pheu Thai Party won 313 seats and the Democrat Party merely netted 152. In fact, the Pheu Thai Party won 265 seats and the Democrats 159.
Sukhumbhand urged the Election Commission of Thailand to probe the legitimacy of all the poll centers so that better polls could be conducted in the future. Suriyasai Katasila, coordinator of the Green Politics Group, tweeted a similar concern to his 12,000 followers.
“I think that politics pollsters should take a break,” said @Suriyasai. “They should hold talks together to look into this lesson and improve their poll conduction to revive their credibility.”
Sing Singkajorn, director of Bansomdej Chaopraya Poll, said that the discrepancy between the exit polls and the election results was due to the fact that many voters might have turned up at polling stations after volunteers wrapped up their data collection at 1:30pm.
Noppadon Kannikar, director of ABAC poll, insisted that his surveying process met with international standards. He urged people, somewhat ironically, not to trust pollsters and promised that he would not make poll results public if they could leave some candidates at a disadvantage.
But while scholars, politicians and the Election Commission are investigating possible slips during the data collection process, some people have pointed to the core of the Thai culture as the main problem. Thai people are reluctant to talk about politics, religion or the monarchy with total strangers, a cultural proclivity that makes accurate polling a difficult task.
