By RK Shih/KHT Staff. AI illustration image.
KAOHSIUNG — A new online poll showing DPP Kaohsiung mayoral candidate Lai Jui-lung (賴瑞隆) barely ahead of KMT candidate Ko Chih-en (柯志恩) has renewed debate over why the race appears close in one of the DPP’s strongest cities.
The survey released by Big Media, or 鉅聞天下, put Lai at 47.78% and Ko at 46.14%, a gap of 1.64 percentage points.
Pearson Data (皮爾森數據) conducted the online survey from July 27 to Aug. 1 among 1,608 Kaohsiung internet users aged 20 or older. The pollster reported a sampling error of plus or minus 2.40 percentage points at a 95% confidence level.
The result was within that stated range and does not establish which candidate was ahead. Because the survey sampled internet users, its reported sampling error also does not account for every possible source of bias.
Other polls have produced substantially different results.
A closed online ETtoday survey released in June put Ko at 46.8% and Lai at 42%. By contrast, a July telephone poll commissioned by Z.Media put Lai ahead 39.6% to 31.6%, while an ERA News telephone survey showed Lai leading 36.2% to 23.6%.
The differing survey methods, sponsors, questions and sample structures mean the results should not be treated as a single trend line. They show considerable uncertainty about the race, not a settled statistical tie.
Still, the Big Media poll points to a question that has followed Lai’s campaign for months: Why has he not built a more comfortable lead in a city where the DPP holds the mayor’s office, all eight district legislative seats and a strong local organization?
One answer may be Ko’s performance among voters outside the DPP base.
The Big Media survey found 41.43% of respondents identified with the DPP, compared with 25.19% for the KMT and 4.93% for the TPP. Another 23.14% described themselves as neutral or as voters who choose candidates rather than parties.
Among that neutral group, Ko led Lai 59.79% to 24.58%. Ko also received 95.02% support among the poll’s relatively small group of TPP-identifying respondents.
The Z.Media telephone poll found a similar pattern. Among respondents without a party preference, Ko led Lai 24.3% to 12.2%. Ko also led among TPP supporters, although Lai remained eight points ahead citywide.
The Big Media poll also examined negative party identification. Its detailed results found that respondents who disliked only the DPP accounted for 39.63% of the sample, the largest of eight individual categories. Of those respondents, 97.55% supported Ko.
However, the same analysis found that the broader anti-KMT bloc remained larger than the anti-DPP bloc, 50.27% to 42.39%. The findings therefore point to a highly consolidated anti-DPP vote behind Ko, but not to an overall electorate that has turned against the green camp.
Another problem for Lai is succession.
Mayor Chen Chi-mai (陳其邁) remains popular, but approval of Chen has not automatically transferred to support for Lai.
The Big Media survey put Chen’s approval at 63.32%, about 15.5 points higher than Lai’s support. Among respondents who were “very satisfied” with Chen, 92.39% supported Lai. But among those who were only “satisfied,” 29.82% supported Lai while 58.23% backed Ko.
The July Z.Media poll produced a similar warning. It put Chen’s approval at 78.5%, while a Mirror Media analysis reported that only 49.3% of respondents satisfied with Chen supported Lai.
That does not mean Chen’s popularity has no value to Lai. It means voters who approve of the current mayor do not necessarily view Lai as the automatic heir to his record.
DPP figures strongly dispute the argument that Ko has caught up.
They have accused the KMT side of promoting a “50-50 race” narrative to build morale and create an impression of momentum. They have also questioned the sponsors and methods behind some of the closer results.
Those concerns do not apply equally to every poll.
A June TVBS telephone poll showing Lai ahead by only 0.6 percentage points was commissioned by Ko, who defended the decision by saying candidates routinely hire professional polling organizations.
The latest Big Media poll was not disclosed as a Ko campaign survey. Its methodology statement says Big Media commissioned and financed the research, with Pearson Data conducting it. No public evidence has been presented showing that Ko or the KMT paid for the poll.
Before its release, Kaohsiung City Councilor Chang Po-yang (張博洋) said he expected another survey showing a narrow race. He argued that such numbers could be used to steady blue-camp morale after Lai established a large support organization involving agricultural, fisheries and irrigation groups, including networks traditionally considered KMT-leaning.
Ko’s campaign rejected that argument, saying no candidate would manipulate a poll that still showed them behind. It also said Lai had established numerous support groups without seeing a corresponding rise in his polling numbers. UDN reported both sides’ responses.
For now, the evidence supports two conclusions.
Kaohsiung remains a DPP-leaning city, and most publicly released polls still show Lai ahead. But Ko appears capable of consolidating KMT and TPP supporters while competing strongly for voters without a firm party preference.
What the evidence does not establish is the size of the gap.
For Lai, the challenge is to consolidate DPP voters, inherit more of Chen’s governing approval and improve his performance among non-aligned voters.
For Ko, the challenge is to prove that the support seen in close or Ko-leading polls, particularly online surveys, can become votes on Nov. 28.
The race may be closer than the DPP would like, but the available polling does not yet prove that Kaohsiung is tied or preparing to change parties.
It does show that succeeding Chen Chi-mai will not be an automatic handoff to the DPP nominee.
