Donovan’s Deep Dives: Reading an election through a polling fog

Sparse surveys and conflicting results make it difficult to gauge whether enthusiasm for Puma Shen is lifting DPP candidates beyond Taipei

• By Courtney Donovan Smith 石東文 / Staff columnist

• • Add TT as Preferred Source

  • 

  • 

  • 

The polling situation this election cycle is dire. The quantity of polls is way down and many conducted are effectively useless. One Oct. 29-30 poll for the Taichung mayor election shows the Chinese Nationalist Party’s (KMT) candidate Johnny Chiang (江啟臣) up by 10 points, while another Oct. 28-30 poll shows the Democratic Progressive Party’s candidate Ho Hsin-chun (何欣純) up by 11 points, which is what the last two polls in Taichung show.

In a report published on Oct. 8 in Voicetank on Taipei polling, author Wang Hung-en (王宏恩) calculated that only one-fifth the number of polls this election cycle had been released compared to the average of the last three election cycles. Outside Taipei, the situation is even worse.

The central question of this election is whether this cycle is like the “Han wave” year of 2018, which saw the popularity of the KMT’s Kaohsiung mayoral candidate Han Kuo-yu (韓國瑜) turn into a rising tide of support for his party nationally, or if it a more normal cycle fought race-by-race over more local issues.

Democratic Progressive Party candidate for Taipei City mayor, Puma Shen, on Nov. 8 attends the opening ceremony of a joint campaign headquarters with fellow DPP-nominated city councilor candidate Wendy Ma in Taipei’s Zhongzheng District.

Photo: George Tsorng, Taipei Times

Local media have quoted DPP insiders as estimating that the popularity of their Taipei mayoral candidate Puma Shen’s (沈伯洋) has led to an “ocean wave” movement that will boost their party by roughly six points nationally. The KMT insists that the excitement over Shen is limited to the same pro-DPP base that made up the Bluebird and Great Recall movements, but is not spilling over into the general public. That would make this cycle a race-by-race contest that the KMT could do well in.

Yet this time, poor polling is obscuring more than it is illuminating which side’s analysis is correct.

BUT WHY?

Wang’s analysis of Taipei polling finds the decline is across partisan lines, and theorizes that much polling is being withheld for strategic reasons. Political scientist Nathan Batto, writing on his blog Frozen Garlic, sums it up this way: “It isn’t that polls aren’t being conducted; it’s that parties and campaigns are choosing to keep the results to themselves for various strategic reasons.”

Betto goes on to add: “For campaigns doing worse than expected, it doesn’t demoralize supporters. For campaigns in the lead, it keeps voters a little anxious so they will be sure to vote.”

That explains the partisan polling, but there is another problem. Some formerly prolific pollsters are gone. For example, RW News is closed, as is the Taiwan Public Opinion Foundation after its former chair, Michael You (游盈隆), was appointed to head the Central Election Commission (CEC).

Then there is the Formosa debacle. Earlier this year there was a split in their leadership and Tai Li-an (戴立安) left. He has organized polls for other outfits, but all controversially skew heavily towards the DPP. One of his polls showed a 13 point lead in New Taipei City, but DPP insiders quoted in local media suggested a lead only in the single digits. It could be Tai is a genius picking up signals others are missing, or he is a partisan hack. For now, I am ignoring his polls until there is more of a track record.

The Formosa poll in August showed a sharp drop in support for both the KMT and the Taiwan People’s Party (TPP). Shockingly, after using the same question for years, they changed the wording — effectively making the results useless for comparison.

LITTLE TO WORK WITH

In the 2024 presidential race, inspired by Batto’s earlier efforts, I created a weighted poll of polls. To test my results, a reader apportioned the undecided results according to the final results, showing that — assuming they voted proportionally — the result of my poll of polls was accurate within one percentage point.

To construct it, I eliminated all highly partisan polling, then gave weighted ranks of low, average and high depending on the reliability of the pollster.

To analyze the possible impact of Shen’s “ocean wave,” I only looked at polls taken after Oct. 1 (when it was clear the wave was real) and only examined results from pollsters I weighted as average or high.

This left only nine polls for the entire country: Two for Taipei, New Taipei City and Keelung, and one each for Kaohsiung, Changhua and Hualien. Worse, the last two are effectively useless for analyzing the “ocean wave;” in Hualien there is no DPP candidate and in Changhua the DPP side is split.

Looking at the new dataset starting from August, preference for all three parties has risen by a few points, but nothing suggesting a six point surge for the DPP — though if support for the president is counted as a proxy, it is way up since the summer (they did not change those questions).

In Kaohsiung, a United Daily News (UDN) poll shows a seven point lead for the DPP’s Lai Jui-lung (賴瑞隆) over the KMT’s Ko Chih-en (柯志恩). That lead is low in this DPP stronghold.

In New Taipei City, the UDN poll has the DPP’s Su Chiao-hui (蘇巧慧) leading the KMT’s Hammer Lee (李四川) 42 to 39, while a Mirror News poll has Su ahead of Lee by 47.7 to 32.8. However, Su’s surge predated the “ocean wave,” so it is hard to tell how much impact it has had.

Keelung is tight. An Apollo-conducted poll shows the KMT incumbent George Hsieh (謝國樑) leading the DPP’s Tong Tzu-wei (童子瑋) by half a percentage point, while a Mirror News poll has Tong leading by 0.7 percentage points.

With no useful polling in most of the country, only one poll in the south and nothing in central Taiwan, it is hard to gauge if the “ocean wave” is having the impact the DPP hopes it will.

The Formosa national polling and the Kaohsiung poll suggest that if it is real, it is less than six percent. However, the Keelung and New Taipei City polling suggests it could be very real.

It is possible that the “ocean wave” is helping a little nationally, but is impacting the commuter regions surrounding Taipei more. One DPP insider estimated that half of supporters cheering on Shen’s campaign in the streets do not live in Taipei City. Likely, they live in New Taipei City or Keelung.

Ultimately, elections come down to turnout. It is possible that enthusiasm for Shen pushes more to the ballot box, even if it does not extend beyond the base as the DPP hopes.

Donovan’s Deep Dives is a regular column by Courtney Donovan Smith (石東文) who writes in-depth analysis on everything about Taiwan’s political scene and geopolitics. Donovan is also the central Taiwan correspondent at ICRT FM100 Radio News, co-publisher of Compass Magazine, co-founder Taiwan Report (report.tw) and former chair of the Taichung American Chamber of Commerce. Follow him on X: @donovan_smith.