- Published on
From 2022, during better times for the Taiwan People's Party -- and before the Taipei City election ordinance banned most campaign billboards.
What's This?
In the run-up to Taiwan's 2022 local elections, I put together a set of previews on the local mayor's races to help me track the campaign. I'm going to do that again here for 2026 and roll these out over the next few days. Some preliminaries follow...
Taiwan's local elections will take place on November 28, 2026. There are contests for over 10,000 different offices, ranging from the mayor of Taipei all the way down to village chiefs in indigenous communities in rural Taitung. In Taiwan these are often called the "9-in-1 elections" (九合一選舉) because they combine voting for nine different types of offices, most of which used to be held at different times, in one single campaign and election day held once every four years. These are:
- (1) Mayors and (2) city councilors in centrally-administered municipalities (直轄市市長,市議員) for Taipei, Kaohsiung, New Taipei City, Taichung, Tainan, and Taoyuan;
- (3) Executives and (4) councilors in counties and county-level municipalities (縣/省轄市長,縣/市議員);
- (5) Township and town heads and (6) councilors (鄉/鎮長,鄉/鎮議員);
- (7) Village and ward heads (村/里長);
- (8) Indigenous "self-governing" district heads and (9) representatives (自治區長,區代表).
This consolidation of election terms and cycles has turned local elections into something like a U.S. federal midterm election. It's obviously a big deal in local politics -- all of Taiwan's elected offices except for the president and the legislature are up for grabs. But it's also a big deal for all three parties that hold seats in the legislature right now, and there are some important if indirect implications for the 2028 races and for the current national political stalemate.
What's at Stake for the Political Parties?
For the KMT, the results will affect Cheng Li-wun's (鄭麗文) influence in the party. As party chairwoman, one of the most important jobs is to manage nominations and coordinate what is increasingly a nationalized election campaign. If the party does well overall, she's well-positioned to steer nominations for the KMT's 2028 presidential and legislative candidates. If it does poorly, there will be mounting pressure on her to step aside, and her influence is likely to decline regardless of whether she does.
For the DPP, the elections will determine whether President Lai Ching-te (賴清德) can hold onto the DPP party chair, and perhaps even whether he'll face an internal party challenger for renomination next year. The precedent here is 2018; then, Tsai Ing-wen had to resign after a terrible showing for the party and then got challenged for the DPP's presidential nomination by...Lai Ching-te, who argued that Tsai was headed for defeat if she was renominated. She beat that back and ended up crushing Han Kuo-yu in the 2020 election, but it wasn't easy. Lai could have to deal with a similar threat if the DPP struggles again in the local races.
And for the TPP, these elections are downright existential. The party peaked in January 2024, when Ko Wen-je (柯文哲) won 26.5% of the presidential vote and the TPP captured eight critical seats to make them the kingmakers in the LY. But since then they've played a strong hand remarkably badly. They didn't win either of the Legislative Yuan leadership positions despite having the leverage to do so, their former chairman Ko is now appealing a 17-year prison sentence for corruption and banned from running in 2028, their new party chair Huang Kuo-chang (黃國昌) lost the KMT-TPP joint nomination for New Taipei to a colorless KMT technocrat, Ko's former deputy mayor Huang Shan-shan is no longer in the legislature, and they had to expel legislator Li Zhenxiu (李貞秀) for failing to renounce her PRC citizenship. From a really promising starting point, they've done little in public to distinguish themselves from the KMT over the last 2+ years, and they're now the most disliked of the three major parties.
So, these elections loom as a crucial moment for the party: they need to win a couple mayor's races and increase their councilor seats (they won only 14 of 910 in 2022) to demonstrate they are still a viable alternative to the two big parties. But the signs so far aren't great. The party entered into a joint nomination process with the KMT and got very little in return: only in Chiayi County (a deep green stronghold) did the KMT back a TPP candidate rather than the other way around. There is also Hsinchu City, where the incumbent mayor Kao Hung-an (高虹安) is running for re-election as an independent, and the KMT party center has decided not to nominate a candidate there. Kao was a TPP member when she was elected, but got kicked out of the party after being charged with corruption; she beat the charges and has not rejoined the party, so the TPP can't really count her as one of their own anymore.
Without any competitive mayoral candidates of its own, the TPP needs a really strong showing in the councilor races to dispel the growing impression they're in (perhaps terminal) decline -- especially important for a party that doesn't have many (any?) consistent values or distinct policy goals. Since the TPP's founding in 2019, it has attracted ambitious candidates and staff mostly because it offered a more promising path to power than waiting behind better-connected people in the DPP and KMT. When people stop believing that the TPP will be the winning team in the future, there aren't many people who will stick with the party because they sincerely believe in its mission...or even know what that mission is.
There is clearly still significant electoral demand for a centrist third party that will keep the DPP and KMT honest and push a reform agenda. But the TPP has not done much to actually deliver on that agenda, despite the extraordinary leverage it currently has in the LY. It's hard for me to escape the conclusion the party has peaked and is headed toward the electoral graveyard where third parties of the past are buried. We'll see if they can prove the doubters wrong this election.
The Electoral Rules
Demonstrators call on the legislature to pass an absentee voting bill.
The Campaign Rules
Elections in Taiwan are handled by the Central Election Commission following the Public Officials Election and Recall Act. Some key dates for the 2026 elections:
- August 20, 2026: CEC issues the formal election announcement
- August 31: candidate registration opens
- September 4: candidate registration closes
- October 23: candidate ballot order determined by drawing lots
- November 8: voter rolls finalized
- November 13: official campaign period for the special municipality mayors races begins
- November 18: official campaign period for county/city mayor, councilors, township heads, indigenous district chiefs begins. Polling blackout begins.
- November 23: campaign period for local representative (ward/village head, township reps, indigenous district reps) begins
- November 28: Election Day
A couple features to note about the campaign process. First, despite the official campaign period being only 15 days long for mayors, and even shorter for the other offices, restrictions on campaigning aren't really enforced in practice. Taiwan has pretty liberal free speech and advertising regulations, and candidates have been putting up billboards and out campaigning for months already.
Second, there's a weird twist in Taipei, which in 2022 passed its own ordinance limiting campaign billboards anywhere in the city. This rule is quite a bit stricter than the national standard and moves Taipei incrementally closer to the tight electoral regulatory regimes in place in Japan and especially Korea. (That's not a good thing.) The Taipei ordinance:
- permits campaign advertisements only on official public campaign boards, party and campaign offices, and vehicles;
- defines a campaign advertisement as a poster, banner or billboard having at least three of the following five items: name, image, ballot number, campaign item, and slogan or trademark; and
- applies to everyone, including private individuals not employed by any candidate or campaign.
But it's also full of loopholes -- in 2024 it did little to limit campaign advertising for the 2024 elections. An obvious way to skirt it is to include only two of the five items on any given poster. There is also lax enforcement -- it requires someone to make a complaint before the city government will investigate. And it's also vulnerable to a constitutional challenge (if and when the constitutional court is functional again), not least since it restricts all displays of campaign posters, even on one's own property. (That's even stricter than Japan's rules.) Nevertheless, this is something to watch this election cycle, as a huge challenge for candidates is to get their names out to uninformed and disengaged voters, and billboards/flags/posters have long been a key part of this effort. This is the first time local elections have been held under this new rule, so it's untested, and advertising is much more important in the low-salience council and ward races than in the legislative and presidential contests.
Third, Taiwan has a polling blackout period that is strictly enforced: no one can publish or share polls from 10 days before the election through Election Day. Taiwan is a poll-crazy place, and this restriction is intended to prevent last-minute polls from affecting voting behavior. The CEC has heavy financial penalties it can impose on violators, and the blackout is generally respected. We'll see if that remains true this cycle.
Fourth, there will be at least three referendums held alongside the local elections, on "introducing caning as a criminal punishment, abolishing the government’s 'nuclear-free homeland' policy, and using traffic fines to improve road safety." At one point the KMT and TPP were discussing using their legislative majority to refer up to 10 different referendum questions to the voters. Since the legislature also changed the law back to require referendums to be held the same day as the local elections -- reversing a DPP change introduced after a voting fiasco in 2018, which itself was a consequence of a previous DPP-enacted change to the referendum law -- there is a danger that the voting and counting process will get bogged down again. The current CEC chairman You Ying-lung (游盈隆) has pleaded with the LY for several months to cut down the number of questions to make the process run more smoothly. And today, they did.
And finally, I'll just note that Taiwan still does not have an absentee or mail voting process, nor does it have early voting. This puts a disproportionate burden on young people, active-duty military personnel, and anyone else who isn't living at their official household to return there to vote. It's been a long-standing complaint that has never been addressed, even during COVID.
The Electoral System
For the mayor, township head, and village/ward chief elections, the winner is decided by simple plurality. But for council and township and indigenous district representative seats, Taiwan still uses the old single non-transferable vote system (SNTV) in multi-member districts. This system has an important implication for partisan competition: the electoral threshold of exclusion is much lower than in the executive races, and small parties can be competitive for a seat by winning 10% of the vote or even less in some districts. That means the usual KMT-DPP duopoly in mayor's races is weakened in local races, and small parties face a more forgiving electoral environment than in elections to the Legislative Yuan.
Traditionally, independent candidates have also fared well in local councilor races. Another truism of Taiwan politics is that the lower down one goes in the hierarchy, the less partisan affiliation matters, so that most village and ward chiefs (~90% in the last election) register as independents.
Finally, it's worth noting that Taiwan has long had a gender quota for SNTV races. Since 1999 it's been a 25% requirement -- i.e. at least one of every four seats in a district must go to a woman. If one of the top four vote-getters is already a woman, nothing changes; if none is, the fourth-place male candidate is displaced by the top female candidate. This rule only kicks in if the district magnitude is at least four, which in practice means it's applied somewhat unevenly across Taiwan's localities. Some counties elect most of their representatives from districts with three seats or fewer; for instance, in Penghu only the Magong City district elects more than four councilors and is therefore covered by the quota. In practice, the actual share of women winning elections to councils is significantly above the 1/4 requirement: in 2014, it was over 30% across all councils, and it's risen from there to about 40% (special municipalities) and 36% (city and county councils), respectively, in 2022.
Importantly, this quota has contributed over the long term to the impressive representation of women at the highest elected offices: over 40 percent in the current legislature. Huang Chang-ling and Nathan Batto both have excellent papers on this topic.
To come: Part 2, Part 3, Part 4, and Part 5.