Repeated observation helps distinguish real shifts among the same respondents from changes caused by a new sample.
Taiwan Political Intelligence
Built to follow how the same people change.
SPI places longitudinal survey answers and observed digital behaviour on the same timeline to study changes in political choice, message response and attention in Taiwan.
The first six-city wave was published in August 2026. The second-wave same-respondent read is now published in Traditional Chinese: the uncommitted share edged down, while women and working-age voters remain the key audiences to understand. This is aggregate research evidence, not an election forecast or an individual voter-targeting product.
Why Taiwan matters
A single poll is a photograph. Movement explains the decision environment.
Taiwan’s elections provide a visible test of party realignment, issue competition and the media environment—with implications for regional policy and Taiwan-exposed organisations.
Issue priorities and party movement help overseas readers interpret the domestic setting without turning sample evidence into a national forecast.
For Taiwan-exposed organisations, political and media shifts can inform the questions to monitor—not predict a specific regulatory outcome.
Institutional context
Taiwan’s local elections select mayors and local representatives. The current SPI work follows early positioning ahead of the 28 November 2026 local elections.
Reading party labels
DPP refers to the Democratic Progressive Party; KMT to the Kuomintang. Candidate names and municipality names follow a consistent English glossary across the localized series.
Published insights
Six studies, six decisions
Each article starts with a practical question and carries its own sample, method and limitations.
Six-city election preview: Taipei’s same-respondent gap narrows again; undecided voters will decide the mayoral races
SPI twice-monthly election brief | A first look at current support shares and changes among the same respondents. Full analysis will follow after the next survey is completed.
Read the election previewTaipei’s gap narrows; Kaohsiung remains close: watch how undecided voters choose next
In Taipei, Puma Shen has narrowed the gap with Chiang Wan-an. In Kaohsiung, Ko Chih-en’s support has risen, bringing her close to Lai Jui-lung. ShareParty Insight’s third survey wave follows the same respondents and finds two different patterns: the number of undecided respondents has fallen more in Taipei, while nearly three in ten in Kaohsiung still have not decided.
Read the Wave 3 articleVoters have not automatically rallied: what changed from August to September
SPI returned to the same six-city voters to examine who remains open, which pressures matter and where the next message test should begin.
Read the English articleUndecided voters are the largest pool; the digital contest has barely begun
Undecided respondents formed the largest group in the linked six-city sample, while observed political advertising remained sparse. The first creative test should answer what is worth saying.
Read the article35.6% undecided: Taiwan’s first six-city election-strategy wave
New Taipei and Kaohsiung were effectively tied, while every city contained a large undecided pool. These are early August signals, not November forecasts.
Read the articleSame issue, different frame: what 1,003 responses reveal
A comparison of 12 policy messages separates the frames to develop first, those needing more evidence and those not yet ready to lead.
Read the articleChallenging an incumbent: continue, improve or confront?
The relative direction differs by challenger. The larger signal is that many respondents would not move simply because of how a candidate discusses the incumbent.
Read the articleLongitudinal by design
Two forms of memory, held on the same panel
Surveys record what people say. Passive observation records what they did. The useful signal is how each changes—and where the two diverge—over time.
Survey memory
What people say
Repeated choices, certainty, issue priorities and voting intention from the same consented panel members.
Action memory
What people do
Consented app, web, media and advertising observations organized into time-based, comparable behavioural signals.
SPI provides research findings at an appropriate aggregate level. It does not provide personal profiles, raw data or voter lists. Observed activity indicates use or attention; it does not by itself prove persuasion, purchase or voting causality.
Stay current. Go deeper when needed.
Start with the latest Taiwan political intelligence—then commission only what your decision requires.
Subscribe for new articles, or contact SPI for a private discussion when you have a defined issue, audience or deadline.
Taiwan Election Brief
A concise reading of published evidence for a defined issue, audience, municipality or decision.
International Tracker
Wave-based monitoring of choice, certainty, issues and attention among the same panel members.
Virtual Slice
A client-specific aggregate research view built around a defined question; raw and person-level data remain with SPI.
Method and limits: Results come from SPI’s consented online research sample and apply to the sample and period stated in each article; they should not be read as direct estimates of all Taiwan voters. Every figure, base and limitation is governed by its source article. App, web and media activity represents observed use or attention. See Data Use and the technical notes in each article.
Stay current—or request a confidential Taiwan briefing.
Subscribe for new articles, or contact Tatt Chen at SPI when you have a defined issue, audience or deadline.