Summing Up the Races

Going into this election cycle, the KMT holds 16 of 22 localities, the DPP holds 5 (Tainan, Kaohsiung, Chiayi County, Pingtung, and Penghu), and there is one independent (Kao Hung-an in Hsinchu City). As of late August, here are my ratings for the 22 executive races: 

Safe KMT (3): Nantou, Kinmen, Lienchiang

Likely KMT (3): Taipei, Taoyuan, Miaoli,  

Leans KMT (5): Taichung, Keelung, Hsinchu County, Yunlin, Hualien, Taitung, 

Toss-up (2): New Taipei, Penghu

Leans DPP (4): Yilan, Changhua, Chiayi City, Pingtung, 

Likely DPP (3): Tainan, Kaohsiung, Chiayi County 

Safe DPP (0): none  

Leans IND (1): Hsinchu City

 If we put all the leaners and likely races in their respective camps, the DPP would have a net gain, and the KMT a net loss, of two, for a breakdown of 12-7, with one independent (Hsinchu City) and New Taipei and Penghu to be determined by coin flips. So the DPP is well-positioned to improve a little on 2022. Some more general observations follow.

(1) This is a sneaky hard election cycle for the KMT. Ever since Taiwan's local elections were consolidated into a single four-year cycle, the conventional wisdom has been that on aggregate they will turn out a lot like midterm elections in the United States: the party of the president loses seats. That was my starting assumption as well, and the implication is that the DPP would struggle this cycle since President Lai is not all that popular.

But there are several structural factors working against the KMT right now. First, they already control almost every winnable locality, including in places that are reliably green at the national level like Yilan, Yunlin, and Chiayi City. This time around they're playing defense. 

Second, the KMT doesn't have many incumbents running. Incumbency helps a lot at the local level, and the KMT won several of these races in 2022 because they had popular mayors running for re-election. Incumbents tend not to be challenged for the nomination by co-partisans and usually go into the general election with the full backing of the party, whereas open-seat contests often result in intra-party splits, renegade candidates, and factional defections. Most local executives are pretty popular, as well. It's just really hard in Taiwan to defeat an incumbent mayor absent a national partisan wave (which, to be fair, occurred in both 2014 and 2018).   

In 2026, the KMT is defending a lot of open seats; 10, to be exact, in New Taipei, Taichung, Hsinchu County, Yilan, Changhua, Yunlin, Chiayi City, Hualien, Taitung, and Kinmen. By contrast, the DPP is defending only three: Tainan, Kaohsiung, and Chiayi County (although Penghu is technically in there too because of Chen Kuang-fu's health concerns). So the playing field is just less favorable to the KMT now than in 2022, and they could lose control of several localities even if the party's net support across all the races still outdistances the DPP.

Third, going into this cycle, the KMT faced a difficult coordination problem with the TPP. As I argued in a series of posts last year, I think these are make-or-break elections for the TPP's future. They need to demonstrate electoral appeal independent of the KMT, and they desperately need to win some local mayor's races  in order to improve their party brand, attract and cultivate new talent into the party, and convince pan-blue voters that they are a viable alternative to the KMT. And yet...

(2) TPP-KMT cooperation seems to be holding well. Surprisingly well! Given the importance to the TPP's future of winning offices this November, and the spectacular failure of the presidential joint-nomination strategy in fall 2023, I expected the party would play hardball with the KMT over nominations. At a minimum, I thought Huang Kuo-chang would demand the support him as the joint nominee in New Taipei in exchange for the TPP dropping bids elsewhere. 

But instead, the TPP has (mostly -- the exception is the Jhubei City race) stuck to their joint-nomination polling process. There are no blue-white splits in any of the county-level executive races this cycle, and the TPP's nominees have consistently stood down when they lost the joint polling primary. This has been especially beneficial to the KMT's chances. Of all 22 localities, the only joint blue-white nominee who's from the TPP is in Chiayi City -- a very difficult race for the blue camp. The KMT has also yielded in Hsinchu City, where the former TPP member Kao Hung-an is running for re-election. And in Chiayi County, both parties are supporting an independent, Wu Pin-jui. In the other 19 localities, the KMT is running a candidate with backing from the TPP.  

So there is much better coordination on nominees on the pan-blue side than I expected. KMT party chairwoman Cheng Li-wun deserves significant credit for working out that deal and getting the TPP to stick with it.

The corollary is that TPP party chairman Huang Kuo-chang is going to deserve a lot of blame for not getting more out of this cooperation agreement. 1 nominee in 22 races is not a great result. And he couldn't even get the KMT to yield in Jhubei City! 

(3) The DPP is well-positioned to improve, at least modestly, on their abysmal result in 2022, which would give President Lai a boost. I can easily imagine a scenario in which the DPP wins New Taipei, flips Yilan, Changhua and Chiayi City, and scores two or three upsets elsewhere: say Keelung, Yunlin, and Hsinchu County. That would be a really good result for the party, and it seems entirely plausible right now. 

The corollary is that Cheng Li-wun's position as KMT chair looks a little shakier than I thought it would when I started this exercise. The scenario above would be quite bad for Cheng -- maybe not resign immediately bad, but enough to erode her influence and dramatically increase the internal party pressure on her. Add in a loss in Hualien or Taitung or somewhere else unexpected, and Cheng probably can't survive as party chair.

On the other hand, a best-case scenario for the KMT (and worst-case for the DPP) would be for the party to win New Taipei in addition to Taichung, Taoyuan and Taipei, hold on to, say, Changhua and Yunlin and win back Penghu, and pull off an upset in Kaohsiung. Then Cheng Li-wun is riding high and strengthens her influence over the party's nominations for 2028. And Lai probably has to resign as DPP party chair and could even face an internal challenge for the presidential nomination. 

(4) The New Taipei race looms large. It's going to shape the narrative of this entire election cycle. If Su Chiao-hui wins there, and the DPP holds on to the localities it runs now and adds pickups in 3-4 others, the narrative will be that the KMT blew it. If Lee Shu-chuan instead wins, it's a very different take-away: the KMT was playing defense and largely held its own.

So New Taipei is the big prize. It's the most populous locality in Taiwan, the DPP has never won it since it became a special municipality, it's got a close partisan balance, and if national issues are going to matter anywhere it will be here. I usually caution against reading too much into the local election results as a harbinger of national elections, but this race may have outsized implications for 2028. 

(5) Issues...what issues? I realize I haven't talked a lot about issues in the campaigns, and instead focused on candidate quality and factional politics. That's because in in local elections there tends to be a split between the big metro races, where partisanship and national trends matter more, and the other city and county races where local ties and factional politics are more often decisive. This cycle looks like it's following that traditional pattern. 

At the central government level, most issues are linked to the interminable fighting between the KMT-TPP controlled Legislative Yuan and the DPP-controlled Executive Yuan. Oversight, nominations, and budgets have been the main areas of contention. On oversight, the TPP especially has been hopping mad that prosecutors arrested Ko Wen-je in 2024, and he's now been tried and convicted. That's made them even less willing to consider engaging with the DPP, and eased their cooperation instead with the KMT. On nominations, the two branches have escalated their confrontations to the point that the Constitutional Court, NCC, and Control Yuan are now all paralyzed. And on budgets, the KMT and TPP changed the revenue allocation formula to distribute a larger share to local governments (mostly controlled by the KMT) and away from the central government. They also delayed review of the FY2026 budget for eight months -- it finally passed in early August.

I don't see a whole lot of evidence so far that the consequences of the national-level stalemate are trickling down into local races in a way that would be decisive. The issues that have broken through in the mayor's races don't map neatly onto the partisan divide: a food safety scandal, transportation and economic development, disaster relief and flood control, and corruption. It's plausible that these could affect some of the metro races. For instance, in Taichung outgoing incumbent Lu Shiow-yen has taken criticism for the food scandal, and in Taipei Chiang Wan-an has been increasingly prominent as a critic of the central government. But in most of the other races these national issues don't look to be very salient. 

What about the China factor? It's there, for sure, but it's constant background noise in local elections. As we know from the electoral whiplash in the 2018-->2020 and 2022-->2024 cycles, the KMT's political liabilities as the China-friendly party matter a lot more when the presidency and control of the LY are at stake than they do in local mayors and councilor elections. 

(6) National Factors Seem Neutral

Right now President Lai's approval rating is about even -- the MyFormosa poll trend above shows him almost dead even as of May, 46-45%. It's notable that he's polling much better than last summer when the KMT won every single recall election. The national anti-incumbent waves of 2014 and 2018 both occurred when the president's popularity was under 30% (Ma Ying-jeou in 2014, Tsai Ing-wen in 2018).  So far Lai looks to be avoiding those lows, and that puts the DPP's candidates in a better position to make their case on their own merits. The party's brand isn't toxic. 

The Election Study Center's partisan ID surveys are now out through June 2026, and they also show a national environment that is not bad for the DPP. Self-identified DPP partisans are up to about 34% of respondents -- that's near an all-time high. The KMT has recovered a bit from their lows in 2022 -- probably by pulling back some pan-blue respondents who were attracted to the TPP for a bit -- but they're still over 10 points behind the DPP. And the TPP is pretty clearly in decline now, falling from 14% in 2023 to under 9% now. Adding up TPP and KMT partisans gets the pan-blue camp to 30%, close to but still behind the DPP. These data suggest a relatively steady state this election cycle: no signs of swelling or collapsing support for either camp. 

And finally, the economy is roaring: Taiwan's annual GDP is estimated to grow at about 11% in 2026. But the gains from that growth are very unequally distributed -- the high-tech sector is doing well on the back of the AI chip boom, while other sectors of the economy remain sluggish, housing costs remain high, and average wages in non-tech sectors remain low. There's not a lot of evidence the DPP is reaping broad political benefits from the tech boom, but neither do I see strong signs the KMT and TPP are capitalizing on economic frustration from those left out. 

See also Part 1 (Introduction), Part 2 (The Big Six), Part 3 (North and Central), Part 4 (South, East, and Offshore Islands)