The Kla Tham Paradox: Party switching and the Incumbency Advantage

by Joel Sawat Selway

The party logo of Kla Tham (Brave Dhamma Party)

Kla Tham (KT) poses an immediate puzzle for anyone trying to make sense of Thailand’s 2026 election. In national polls, the party barely registers, typically drawing less than 2% of intended constituency votes. Yet at the same time, KT is fielding more than forty sitting MPs in constituency races. This tension—between weak party-level polling and a large pool of incumbent candidates—captures a broader problem with election forecasting in Thailand.

In the previous two articles in this series, I showed that translating polling data into constituency seats is far from straightforward. Using national and regional polls, a range of reasonable modeling choices produced a wide range of different seat projections, many of them far removed from expert forecasts. The problem was not simply whether the data were national or regional, nor whether one handled undecided voters one way or another. The deeper issue was that polls observe party preferences, while constituency outcomes are driven by candidates.

The Kla Tham paradox illustrates this dilemma for forecasters. The party’s strategy is not to build support from scratch, but to aggregate sitting MPs and former candidates from across the conservative spectrum. That approach may or may not succeed electorally—but it cannot be evaluated by polling numbers alone. To assess its prospects, we need to understand how incumbents actually perform, how party switching affects their chances, and how those effects vary across different types of constituency races.

In what follows, I show that incumbency is not as big an advantage as most observers might assume. Its advantage, moreover, is not uniformly distributed. Incumbents historically have performed better in some parties more than others; party-switching incumbents do worse on average than those who stay with the party from the previous election; and the competitiveness of the district matters a lot.

Incumbents and Party Switching in 2026

KT is not the only party courting MPs from other parties. Table 1 shows that about half of the incumbents that BJT is running have switched from other parties. Most of the other main parties have very low rates of party switching, but they vary on the number of incumbents they are fielding. PT has the highest number of incumbents running (93), only 5 of which switched from other parties. PP is fielding 84 incumbents, but has no switchers. DP also has no switchers, but is relying heavily on non-incumbents. KT is a unique party in this respect: almost every single one of its 43 incumbents the party is fielding have switched from other parties. This number is likely the basis for expert predictions on KT’s number of seats.

Party (2569)	Total candidates	Incumbent MPs	MP switchers	Former candidates (2566, lost)	Switchers who are former candidates	Total Party switchers (2026)	True newcomers
ประชาชน (PP)	400	84	0	109	26	26 (6.5%)	207 (51.8%)
เพื่อไทย (PT)	398	93	5	162	69	74 (18.6%)	143 (35.9%)
ภูมิใจไทย (BJT)	394	121	64	140	86	150 (38.1%)	133 (33.8%)
ประชาธิปัตย์ (DP)	400	4	1	198	93	94 (23.5%)	198 (49.5%)
กล้าธรรม (KT)	334	43	42	133	133	175 (52.4%)	158 (47.3%)
เศรษฐกิจ (EP)	263	1	1	106	105	106 (40.3%)	156 (59.3%)
รวมไทยสร้างชาติ (UTN)	250	0	0	135	100	100 (40.0%)	115 (46.0%)
พลังประชารัฐ (PPRP)	199	3	2	120	88	90 (45.2%)	76 (38.2%)
ไทยสร้างไทย (TST)	69	1	0	38	18	18 (26.1%)	30 (43.5%)

Table 1. Incumbents and former candidates running in 2026, by party

Incumbent rates only tell us part of the story of parties’ overall “constituency experience” in national elections. A large number of candidates running in 2026 were losing candidates in 2023. The DP leads in this metric: almost half of the 400 candidates it is running were losing candidates in 2023. At the low end of the major parties is PP with just 109 losing candidates from 2023, a little over a quarter of all its candidates. The final column, then, reflects the overall constituency electoral experience. Larger numbers (and percentages) indicate a slate with less electoral experience. From this perspective, PT and BJT are far ahead of the other parties in fielding candidates with prior (recent) electoral experience.

We can also see the overall proportion of a party composed of candidates switching from other parties. KT leads the way on this metric, with 52.4% of its candidates having competed under a different party banner in the last elections. BJT comes second, though EP, UTN, and PPRP are all slightly higher, demonstrating how broad a strategy this is. Notably, both PT and DP are much lower than these other parties, about half the rate of incoming party switchers. Not surprisingly, PP is the lowest.

How well do incumbents do in Thai elections?

The assumption for KT seems to be that because it has recruited 42 incumbents this will be around the number of seats it will eventually win. But how well have incumbents done in recent Thai elections? Table 2 shows how many incumbents parties ran in the 2023 elections, how many of those incumbents won, and the win rate of each party. First, the bottom row in the table shows that overall, 308 of the 400 sitting MPs contested in 2023. However, only 125 of them actually won their seats. For context, in the US the incumbency win rate is around 90% and 70-80% in the UK. Incumbency is truly an advantage in these countries. But in Thailand, at least in recent elections, the rate is much lower.

Party	Incumbents Ran	Incumbents Won	Win Rate (%)
Pheu Thai (PT)	102	45	44.1
Bhumjaithai (BJT)	82	29	35.4
Palang Pracharath (PPRP)	42	15	35.7
United Thai Nation (UTN / RTSC)	31	11	35.5
Democrat Party (DP)	25	7	28.0
Move Forward / People’s Party (MFP / PP)	9	8	88.9
Chart Thai Pattana (CTP)	7	6	85.7
Prachachat Party (PC)	5	4	80.0
Thai Sang Thai (TST)	3	0	0.0
Chart Pattana Kla (CPK)	2	0	0.0
All parties	308	125	40.6%

Table 2. Incumbent wins by party (2023)

The success rate of incumbents also varies by party. Parties that field less than 10 incumbents are a mixed bag. They either have very high success rates (PP, CTP, and PC), or no success at all (TST and CPK)! Of those who fielded 25 or more, the DP was the least successful with just a 28% win rate, while PT was the most successful with 44.1%.

This matters because KT’s incumbents mostly come from the two parties whose incumbent win rate was ~35% in 2023.

A second contextual feature of incumbent success rate is whether a candidate has switched parties or not. Overall, the success rate of incumbents who switched is half that of those who stayed in their parties: 24% to 48%. This is a large switching penalty—and it further varies by party. BJT had the biggest switching penalty: only 10.6% of incumbents who switched to BJT won their seats. Part of the party effect, then, is driven by the fact that some parties are composed of higher numbers of switchers than others.

This has huge implications for KT candidates in particular, almost all of whom switched from other parties.

Party (2023 label)	Stayed (N)	Win rate (stayed)	Switched (N)	Win rate (switched)
Move Forward (MFP)	9	88.9%	0	—
Chart Pattana Kla	2	0.0%	0	—
Chart Thai Pattana (CTP)	7	85.7%	0	—
Prachachat	5	80.0%	0	—
Democrat (DP)	25	28.0%	0	—
Palang Pracharath (PPRP)	37	35.1%	5	40.0%
Bhumjaithai (BJT)	35	68.6%	47	10.6%
Pheu Thai (PT)	92	43.5%	10	50.0%
United Thai Nation (UTN/RTSC)	0	—	31	35.5%
Thai Sang Thai (TST)	0	—	3	0.0%
Total	212	48.1%	96	24.0%

Table 3. Incumbent wins by party and switching status (2023)

The Effect of Race Structure on Incumbency

Lastly, we look at where parties fielded their incumbents. Table 4 shows us that the more competitive the district race, the lower chance an incumbent has of being re-elected. While that may seem obvious, we can see it is even more difficult for party switchers to win in competitive districts: 20.4% vs. 26.0%.

Race structure (ENP bin)	Incumbents ran (total)	Switching incumbents ran	Switcher win rate (%)	Non-switching incumbents ran	Non-switcher win rate (%)	Total incumbent win rate (%)
2-horse (1.51–2.50)	19	2	50.0	17	47.1	47.4
3-horse (2.51–3.50)	205	43	34.9	162	35.8	35.6
>3.50 (4-horse +)	176	49	20.4	127	26.0	24.4
Total	400	94	27.7	306	32.4	31.0

Table 4. Incumbent wins by race structure and switching status (2023)

What’s perhaps more important, however, is that parties differed significantly in the types of races they competed in in 2023. The incumbents of the three major conservative parties (UTN, PPRP, and DP) each had over 50% of their incumbents competing in the most competitive districts (4-horse + races). Should that pattern continue into the 2026 elections, it makes KT’s job even harder.

Party	Race structure (ENP)	Incumbents ran	Incumbents won	Win rate (%)
Move Forward (MFP)	3-horse	4	4	100
	4-horse	5	4	80
Chart Pattana Kla	3-horse	1	0	0
	4-horse	1	0	0
Chart Thai Pattana (CTP)	3-horse	5	5	100
	4-horse	2	1	50
Prachachat Party	2-horse	1	1	100
	3-horse	4	3	75
Democrat Party (DP)	2-horse	1	0	0
	3-horse	11	5	45.5
	4-horse	13	2	15.4
Palang Pracharath (PPRP)	2-horse	4	2	50
	3-horse	16	6	37.5
	4-horse	22	7	31.8
Bhumjaithai (BJT)	2-horse	1	0	0
	3-horse	47	21	44.7
	4-horse	34	8	23.5
United Thai Nation (UTN / RTSC)	3-horse	15	7	46.7
	4-horse	16	4	25
Pheu Thai (PT)	2-horse	8	6	75
	3-horse	55	25	45.5
	4-horse	39	14	35.9
Thai Sang Thai (TST)	4-horse	3	0	0

Table 5. Race structure by party in 2023 election

Mapping KT’s 2026 Race Structure

If incumbency only provides a reliable advantage in limited-competition races, then the key question for 2026 is not how many incumbents a party has, but whether they have switched parties or not, which parties they switched from, and most importantly what kinds of races those incumbents are contesting.

We already know that all but one of the incumbents KT is fielding are switchers, while ~50%  of BJT’s are switchers. Just over half of KT’s switchers come from PPRP, and in the last election most of PPRPs races were in the more competitive 3-horse or 4+-horse races where PPRP’s win rate was only around 35%. In short, KT has inherited a group of incumbents in the most precarious districts. Of the 43 incumbents that KT is fielding, 23 (or 54.8%) of them are competing in what were the most competitive races in the last elections. Thus, despite 42 sitting MPs switching to KT, the majority of them won their seats with only about a quarter of the constituency’s vote, perhaps suggesting a weaker showing for KT than the experts predict.

What about other parties? Only 39.5% of BJT’s incumbents are in the most competitive districts from last election BJT and more than half ran under the BJT in last elections. PT is best poised for its incumbents to do well. All but 5 of the 93 incumbents they are running are existing PT MPs, and they have the fewest incumbents competing in the most competitive districts (31.5%).

PP has no incumbents that switched from other parties, so it does well on that front, but it does have the highest proportion of incumbents competing in the most competitive races: 56.5%. However, PP’s performance in the polls makes it best poised to win in tight races.

Speaking to the Experts

This analysis has highlighted the importance of incumbents, party switching, and constituency race structure. The results make me seem less convinced that KT can reach 40 seats. However, I am more convinced they can win around half of those. Having incumbents on a party’s slate can be good, but KTs incumbents are of the least ideal variety: party switchers in highly competitive constituencies. The incumbency win ratio is around 40% on average, and KT would seem to be under that average. 10-15 seats thus seems the most likely prediction. 20 seats would be considered a good showing. For 40 seats, just about every other uncertainty has to fall in their favor. I cannot see a scenario in which they win 70 seats.

Both the polling analysis from the first two articles, plus this incumbency analysis leads me to believe the experts are right about BJT. 140 was the number that kept coming up repeatedly in the models transforming votes to seats. And here, we see that BJT has by far the most incumbents on its slate. While roughly half of them switched from other parties (generally less ideal), they face a good race structure on average. I would expect their incumbents to do better than average, putting them on solid ground.

The polls analysis is extremely favorable to PP. However, they have fewer incumbents running than either PT or BJT, and they are running in a higher proportion of competitive districts. I would expect their incumbency win ratio to be right around the average, all things considered. And while most commentors point to their lead on the party list, they will need to perform well in the constituency races too. Last time out, PP surprised the experts, But if they are going to repeat their success they will have to rely the most heavily of all the parties on true newcomers, which make up over 50% of their candidates.

Lessons for Future Forecasting

There is an increase in the number of incumbents running this year. In all, 350 of the 400 constituencies will include an incumbent. These are the types of candidates we know the most about and can model accordingly.

We also have information on other types of candidates, including losing candidates from last round, whether they switched parties, and how close they were to victory.

All this suggests that we can turn incumbency and party switching into forecasting inputs. If we know who moved, whether they were incumbents, previous candidates that lost, or new challengers is important information that we can combine with polling data to improve our predictions. In addition, constituency-level inputs such as race structure must be incorporated. The goal is to make the prediction of Thai elections more transparent and more data-driven.