Forty-four percent of 2,011 people interviewed by the United Daily News said they would vote for Ma...Yes, that's right. AFP ran a piece on the polls from this election, used only polls from pro-Ma newspapers, and didn't mention to its readers (the poor things) that the polls it was using were from papers supporting Ma.
A TVBS news channel poll showed similar results, as Ma maintained a lead of eight percentage points with 45 percent over...
However, Ma and Tsai were much closer in a survey issued by the Taipei-based China Times...
But there is more....
Taiwan polls issued close to presidential elections have a history of predicting the outcomes relatively accurately.In 2008, late Feb polling for the Mar 22th election had the KMT up by 28% (UDN) and 26% (China Times), not "up to 20%". In the China Times March 10 poll, the numbers were an identical 26% (UDN had Ma with a 30% lead according to this). They did predict the winner correctly, but underestimated the DPP numbers by 17-18% (and Ma's by about 10%). They way overestimated the victory margin. I suppose you could call that relative accuracy.....
In 2008, opinion polls indicated that Ma enjoyed up to a 20-point advance over DPP candidate Frank Hsieh, and he went on to win the election by almost 17 percentage points.
It is quite true that in '04 a number of polls, including the last China Times polls, had the two sides much closer. It is difficult to quickly round up poll data from the 2000 election. Niou and Paolino note:
According to a telephone survey conducted by the China Times on January 20, 24% of the respondents were inclined to vote for the independent candidate James Soong, 23% for the Democratic Progressive Party (DPP) candidate Chen Shui-bian, and only 19% for Lien if the election were held then.In 2000 the publicly reported polls were so bad, especially the KMT polls which claimed that Lien had a huge lead over Soong but was neck and neck with Chen, that many KMT voters were misled into thinking that Lien was leading and voted for him even though he had no chance of winning.
This experience may explain the kind of polling that is being reported in the local media for the current election. Apparently many netizens are reporting being called by pollsters who proceed to ask only about Ma and Tsai, and do not mention Soong at all. If these reports are true, by reducing Soong's chances to win, perhaps certain pollers may be hoping that voters will vote strategically for Ma.
ADDED: There is a round-up of current polls here:
To summarize the results, all standard polls show Ma in the lead, with a range of margins ranging from 8 points (United Daily News, TVBS) to 0.7 points (Taiwan ThinkTank). The respective blue and green biases of those three sources seem to show through clearly. The DPP also released its internal polls, which are historically quite accurate, but of course they don’t release the parts they don’t want people to see. With a different methodology consisting of combining polls from over 60 legislative districts, they announced that they expect Tsai to win by 1 point, on a turnout of 78-80% (for more details of the DPP’s methodology, see this report in the Taipei Times.) Finally, the much-talked about xFuture/NCCU Exchange of Future Events has Tsai in the lead by 7.2% (Tsai 49.8, Ma 42.6, Soong 10.7) and this trend has been consistent since mid-December. Exchange of Future Events claims accuracy of 95% two months ahead of 2008 presidential elections and 97.6% on the eve of the election day.The "International Committee for Fair Elections in Taiwan" which operates the website consists largely of pro-Green individuals.
UPDATED: Businessweek (Bloomberg) did the same thing.
[Taiwan] Don't miss the comments below! And check out my blog and its sidebars for events, links to previous posts and picture posts, and scores of links to other Taiwan blogs and forums! Delenda est, baby.