New Taipei Mayor Race: Li Sichen Leads Su Qiaohui by 6.79 Points

The New Taipei mayoral race is heating up as the year-end elections approach. A recent poll showed that Democratic Progressive Party candidate Su Qiaohui was nearly neck-and-neck with her Kuomintang rival, former Taipei Deputy Mayor Li Sichen, with a margin of less than 1 percentage point. However, according to the latest poll by the Juhuo Tianxia News Network, Li Sichen's support rate is 50.85%, leading Su Qiaohui's 44.06% by 6.79 points. The poll results show that Li Sichen's lead is not based on party advantages but on the support of centrist voters and the People's Party. Among the respondents in New Taipei, 33.02% identified with the Democratic Progressive Party and 32.09% with the Kuomintang, with both parties having similar proportions. Among the 23.52% of 'vote for the person, not the party/neutral' centrist voters, Li Sichen led with 54.81% compared to Su Qiaohui's 33.21%. Among those who identify with the People's Party, 81.97% support Li Sichen. Additionally, 78.36% of voters who supported People's Party founder Ko Wen-je in 2024 also chose Li Sichen. The poll results show that Su Qiaohui's advantage is concentrated in the 30-39 age group (53.32% vs. 39.31%) and the 40-49 age group (49.78% vs. 42.55%). Li Sichen, on the other hand, firmly holds the 50 and above age group (59.81% for ages 50-59 and 58.65% for ages 60 and above). However, among young voters aged 20-29, the two are almost tied (Li 45.51%, Su 45.07%), but when looking at gender, young women support Su Qiaohui at 58.30%, while young men support Li Sichen at 54.98%. The survey was conducted by Juhuo Tianxia News Network, commissioned to Pearson Data, from July 6 to July 11, 2026, with a total of 6 days. The effective sample size was 1766, with a sampling error within ±2.30% at a 95% confidence level. The sampling method used an online active survey distribution approach, using a Data Management Platform (DMP) to conduct stratified random sampling by gender, age, and place of residence. The survey was supplemented with user profile tag analysis based on online behavior to ensure the uniqueness of the survey subjects. The accuracy of the respondents' gender, age, and place of residence was verified using online behavior and database tag matching, combined with questionnaire design for double authentication to ensure data accuracy and reliability. Samples that were not eligible for random sampling, such as those obtained through questionnaire forwarding, repeated responses, and other samples that did not meet the random sampling conditions, were not included in the effective sample. Sample representativeness and weighting were conducted using ratio estimation, with population parameters based on the Ministry of the Interior's published data on citizens' age, gender, and household registration as of May 2026, combined with Pearson Data DMP to correct online population characteristics. Continuous corrections were made item by item to ensure that the sample characteristics matched the population structure.