Electoral tendencies of the Czech public – January 2025 – Weekly election model – week 04

The research sample is representative of the Czech adult (18+) population and was created using quota sampling; the quotas regard the participants’ sex, age, education, region of residence, and the population size of their place of residence. We further strengthen the representativeness of the sample by weighting for deviations from the main quota characteristics, as well as weighting for past voting behaviour in the October 2021 Chamber of Deputies election, the January 2023 second round of the presidential election, and for social/employment status. Data collection using online interviewing (CAWI method) of participants registered with the Czech National Panel is provided by the STEM/MARK agency, which, like STEM and the Czech National Panel, is a member of the SIMAR association of research agencies and follow its industry standards.

The fourth election model of the series was based on the responses of 1,549 participants. The respondents were interviewed in 3 waves between the 10^th and 28^th of January 2025.

For the best possible results, we check the attention of respondents when filling out the survey, check the logical link between the answers to questions, and we exclude survey answers that were completed suspiciously quickly. The statistical error of the sample is around ±1 percentage point for the smallest parties, and around ±3 percentage points for the larger parties.

The reported preferences and model are part of a unique research series, The Weekly Model, produced by STEM for CNN Prima News. Detailed description of the methodology can be found on our website:
https://www.stem.cz/?metodologie=tydenni-model-pro-cnn-prima-news-2025

SUMMARY

The STEM research institute, in cooperation with CNN Prima NEWS, provides the public with a new weekly election model.

The election model created by STEM using data from the second half of January continues to be dominated by the ANO movement, which slightly strengthened its position during January to 35% and would secure 89 seats in the Czech lower chamber. From the currently governing parties, the SPOLU coalition is the strongest. The parties that would also enter the Chamber of Deputies include the STAN movement, SPD, the Pirate Party, and likely the STAČILO! movement.

The Motoristé Sobě party has significant potential, but currently only a small group of committed voters, so with a result of around 4%, it would not enter the Chamber. “Support for Přísaha and Motoristé Sobě is still unsettled. Whether it can be stabilized and grow a solid voter base will be shown by future developments,” adds Kratochvíl, noting that Motoristé Sobě show higher support among the youngest age group, among people with secondary education, and also among men.

With a result of 5.8%, as projected by the current model, the STAČILO! movement would win 12 seats. While it does not have as much potential as Motoristé sobě, it relies on a bigger, more stable base of supporters. STAČILO! supporters are mainly found among older age groups and people without a university degree.

In the area of protest votes, the SPD movement continues to dominate. It currently has the support of just under 9%, which would correspond to 19 seats. For the first time in a while, it is approaching its result from the last parliamentary elections. As some smaller parties (Trikolóra, PRO, Přísaha) weaken, some protest votes may once again shift toward SPD. The movement traditionally draws strong support from people without a high school diploma.

In the second half of January, the SPOLU coalition would reach 18.8%, translating to a possible 46 seats. The STAN movement saw a gradual increase during January, reaching 9.4%, which would give it 24 seats. Support for the SPOLU coalition is evenly distributed across gender and age groups, while STAN draws its main voter base from younger voters.

The former fifth coalition partner, the Pirate Party, has stopped its decline, but its support still falls short of previous levels, fluctuating between 6 and 7%. The most problematic aspect remains the party’s weakening among young voters. A significant portion of its former supporters likely shifted to STAN or joined the ranks of undecided voters.

PARTY PREFERENCES

The party preferences show the distribution of political preferences among the entire population of respondents, i.e. including undecideds and people who are unable or unwilling to name a party preference or openly say they will not vote for any party. STEM asks for political preferences using an open-ended question and records the participants’ answers verbatim. Thus, we list groupings mentioned by the participants alongside undecideds and negative responses.

The preferences for each party are lower than would be expected from the actual election result, which is decided only by the voters who actually turn out.

“If you voted, which party would you vote for next week in an election for the Chamber of Deputies of the Parliament of the Czech Republic?”

(data for the second half of January 2025;
open question – no selection from a list of parties or coalitions)

ATTENTION: PARTY PREFERENCES DO NOT REFLECT THE ELECTION RESULT!
 THE TOTAL MAY NOT ADD UP TO 100% DUE TO ROUNDING.

VOTER MOTIVATION

At the latest since the 2013 Czech general elections, we have seen significant changes in the electoral strategy of various voter groups. The number of sworn party voters has declined, the spectrum of political parties and movements has grown, and there are more frequent changes of party preference. The fluctuation between voters and non-voters is intensifying and the proportion of those who decide whether and whom to vote for only shortly before the elections is increasing. The peak of these phenomena so far was the 2021 parliamentary election – six months before the elections, a third of potential voters had been unsure about their participating in the election or who they might vote for.

Thus, uncertainty is becoming increasingly important in estimating political preferences and the fluctuations of electoral sentiment, and it makes sense to look at this phenomenon more closely and analyse its role. We have therefore prepared an overview of the groups of the Czech adult population according to their motivation to vote.

We thus find that the Czech public can be divided into three groups in terms of voting motivation:

1. Less than half (40–50%) of people go to the polls regularly, their participation in the parliamentary elections is very likely and they are mostly clear about whom they want to vote for.
2. One third of the population (around 35%) is either unsure about their participation in the elections, hesitates about their participation or is unclear about their choice of party.
3. A fifth to a quarter of the population is made out of people who do not vote and their motivation to vote is negligible.

People eligible to vote
grouped by motivation to vote in the second half of January 2025 [%]

The sum of the share of Stable voters and Undecided voters (83% in the second half of January) can be considered the theoretical maximum turnout that could be achieved in the Czech Republic currently. The overview of the motivation to vote further serves as a starting point for the construction of the turnout and voting model.

WILLINGNESS TO VOTE FOR INDIVIDUAL PARTIES/MOVEMENTS

In addition to stating direct support for a party or movement in an open-ended question, the research also inquires about willingness to vote for the main Czech political groupings. On a seven-point scale, participants can express the degree of “acceptability” for each of the parties, movements or coalitions.

Willingness to vote for respective political parties and movements
– second half of January 2025
“If the election for the Chamber of Deputies were held today, would you…”

The maximum hypothetical amount of support, the so-called potential for a given grouping, is calculated as the sum of the weights of the votes of those who chose one of the top three answers (“Definitely would vote”, ”Rather would vote” vote and “Maybe would vote”) for this grouping and who also belong to the groups of Stable and Undecided voters.

ELECTION MODEL

It is not easy to build an election model because it is difficult to estimate various parameters: which survey participants will actually participate in the election, whether the voters’ turnout motivation will be similar to the previous election or not, how last-minute emotions will be reflected in the election result and so on.

In surveys, members of the public like to present themselves as responsible citizens: when we pose a direct question, we find that up to 70% of voters want to participate in elections, but the actual turnout does not usually reach that level. Thus, the key to model electoral outcomes and voting patterns is to estimate the behaviour of hesitant and undecided citizens.

We follow a three-step process while calculating the STEM election model.

First, we identify the main groups within the public based on their motivation to vote, as described above. We do not take into account the answers of survey participants identified as Non-voters, and for the remaining two groups (Stable voters and Undecided) we estimate their probability of participation in a hypothetical parliamentary election.

The turnout probability values are used to recalculate the data for the Stable voters group. This yields a new data set, which we refer to as Loyal voters. We treat these as the bottom threshold for the hypothetical support for the respective electoral groupings (parties and movements). In other words, these are the supporters that an electoral grouping is very unlikely to lose in the medium term, and whose support would likely only be shaken by some major scandal. Loyal voters form the basis of the STEM electoral model, as they represent about two-thirds of the votes in the model. With a modelled turnout of 61%, they make up about 42 percentage points. Undecided voters contribute the remaining third of the votes in the model (in the second half of January 2025, this is about 19 percentage points). These votes are weighed not only based on the participant’s probability of turnout but also based on the probability of them voting for the parties they are seriously considering.

Figure of the process of selecting participants for the model

STEM Election Model
– simulated election outcome under a modelled turnout of 61%
Second half of January 2025 (results in %)

Election model, loyal voters, maximal hypothetical vote share
second half of January 2025 (results in %)

SEAT PROJECTION BASED ON THE STEM ELECTION MODEL

We can estimate the number of seats for each political grouping if we apply the legal rules for converting vote shares (as given by the STEM model) into seats in the Chamber of Deputies and estimate voter behaviour patterns in the respective electoral constituencies and the national turnout level.

The seat redistribution system in the Czech Republic traditionally favours parties with the highest share of the vote. This so-called “winner’s bonus” tends to be larger the greater the gap between the leading party and the others. An advantage is also held by parties with evenly distributed support across regions, or by those that perform strongly in the four most populous regions—Moravian-Silesian, South Moravian, Central Bohemian, and Prague—where nearly half (94) of all seats are allocated.

It becomes evident that even slight shifts in model projections—well within the statistical margin of error—can significantly impact the interpretation of the situation. According to the current STEM model, the STAČILO! movement would enter the Chamber of Deputies, while the Motorists for Themselves party would not. It is very likely that the strength of individual groups has not changed significantly compared to previous weeks. Differences in the number of seats won by political parties in various measurements should be seen not only as an illustration of the uncertainty inherent in mathematical models, but also as proof that in elections, every single vote can make a difference.

The ANO movement continues to maintain a clear lead over the second-place SPOLU coalition and other parties in the STEM model. Moreover, it appears to enjoy relatively strong support across all regions. This would likely translate into a solid bonus during seat distribution. On the other hand, the situation is particularly volatile for parties hovering just above the five-percent threshold. Even small changes in vote shares or slight shifts in the relative positions of the three smaller groups can result in notable changes in the number of seats won.

In the case of the STAČILO! movement and the Pirates, the seat allocation rules become apparent—under certain conditions, a party with a lower percentage in the model (in this case, STAČILO! with 5.8%) may end up winning more seats than a party with a higher share of the vote (Pirates with 6.4%). A similar situation occurred in the last election, when the SPOLU coalition won the most votes, but the highest number of seats went to the narrowly second-place ANO movement.

Due to its rules, the Czech seat allocation system is clearly sensitive to even small differences in vote share near the five-percent threshold. It is also sensitive to the number of parties surpassing the 5% mark.

Seat projection in case the STEM election model matched the actual results
– for the second half of January 2025 (200 seats in total)

TRENDS IN THE STEM ELECTION MODEL

STEM election model results since the last (2021) general election [%]

Note: For parties that were ran as members of electoral coalitions in 2021, we calculated the individual parties’ vote share based on a STEM study undertaken shortly after the 2021 election.
**Measured as KSČM until November 2024.
***Until October 2024, the result of the SPOLU coalition was measured as the sum of the coalition’s three constituent parties.

ELECTION MODEL RESULTS BY SELECTED INDICATORS

We can examine voter behaviour in the election model grouped by generic socio-demographics. It is good to keep in mind that when splitting the election model into demographics, the statistical error also increases, and thus these sub-models should be interpreted with caution.

Election model split by sociodemographic categories
Second half of January 2025 [%]

OVERLAPS BETWEEN THE POTENTIAL VOTER BASES OF THE SELECTED PARTIES AND MOVEMENTS

The model shows the current support for the parties, which is made out of, to varying degrees, by loyal and undecided voters. To better understand what is happening in politics, it is useful to assess the relative proximity of the parties, i.e. where each party can potentially gain additional support from.

Note: The number under the party name given in the column represents the party’s electoral potential. The numbers in the cells are the shares of participants who belong to the potential of both the party given in the row and the party given in the column.

MAIN FINDINGS

1. STEM is publishing a weekly election model constructed using a methodology that takes into account the increased level of uncertainty and people’s hesitancy to vote.
2. Around one-fifth of the public are non-voters, while two-fifths to one-half usually vote and are usually clear about whom they will support. The remaining third are unsure about coming to the polls or unclear about their choice of party. It is this group and the estimation of their voting behaviour that is crucial for voting patterns and the electoral outcome.
3. The model captures the current sentiments among voters and the possible distribution of power that would result from a hypothetical election to the Chamber of Deputies if it were held in the second half of January. The parties and movements are slowly starting up their election campaigns and specify their intentions. Some have already announced joint election efforts, and further mergers or alliances may still occur.
4. The model shows the current support of the parties, which is made out of, to varying degrees by, loyal and undecided voters.
5. Take note: the election model does not predict how the actual election will turn out but rather captures the current sentiments and estimates how the election might turn out if it were held now. The election year has always been a dynamic one, and there have been large changes in party preferences in the run-up to all recent general elections.
6. The Czech political scene continues to be dominated by the ANO movement, which is the only electoral grouping to stably achieve results above 30% in the election models, including the current one. In STEM’s weekly model, the movement has been gradually strengthening since the beginning of the year. It remains unchallenged as the leading voice of the opposition, with a broad base of loyal voters. The ANO movement finds strong support among citizens without a high school diploma and among people over the age of 60. Women are more likely to vote for it than men. However, it does not perform poorly in any specific sociodemographic group.
7. The SPOLU has the support of nearly 19 % of voters. This is a slight decrease of support, in comparison to last year’s model by STEM. However, this is still the highest recorded support for SPOLU in January 2025. Support for the SPOLU coalition is even across gender and age groups.
8. The STAN movement, as the remaining member of the governing coalition, has managed to maintain an elevated level of support compared to the past – in the STEM model, it holds support around 11%. The movement likely still partially benefits from the decline in support for the Pirates and from its success in regional elections. STAN’s core voter base is found among younger and youngest voters.
9. Government parties and the Pirates perform well among people with a university or secondary education, but their position is weaker among those without a high school diploma.
10. The governing coalition parties and the Pirates are, in total, performing worse than they did in the last parliamentary elections. However, in January there is no clear trend of voters from the original five-party coalition moving systematically to other groups
11. The SPD movement continues to compete for the role of the main representative of protest votes. In the current STEM model, it has nearly 9% support and is thus approaching its result from the last parliamentary election. As some smaller parties (Trikolóra, PRO, Přísaha) weaken, a portion of protest votes may once again shift toward SPD. The movement has traditionally been strong among people without a high school diploma.
12. The Pirates partially consolidated their support in January compared to the fall of last year, though the weakening from earlier crises persists. Compared to early January, they slightly declined and are currently polling above 6%. The most problematic area remains their decline among young voters.
13. The STAČILO! movement hovers around the five-percent threshold in STEM’s January waves, and would likely enter the Chamber of Deputies at this point. Its supporters are primarily found among older age groups and among people without a high school diploma.
14. In the fourth January wave of the STEM model, the Motoristé sobě party repeated its result of around 4%. It has high and still growing potential. Public awareness of the party is clearly increasing, and a group of people willing to vote for it is being formed. However, it still lacks a strong, committed core of supporters (though even that share is slowly increasing). To achieve more stable results above the 5% threshold, it will need to build a firmer base. The party shows higher levels of support among the youngest age group and also among men.
15. Both the SOCDEM party and the Přísaha movement have seen a slight decline in support. According to current STEM models, neither would likely enter the Chamber of Deputies. SOCDEM has long struggled to re-engage its former supporters or attract new ones. Přísaha has failed to build on its solid result from last year’s European Parliament elections.
16. For government or Pirate Party supporters, the Green Party has long been a second or third choice, which results in only low actual support and a weak voter base. It remains unclear whether it could surpass its 2021 result in a real election
17. The potential margin of error is +/- 1 percentage point for smaller parties and +/- 3.0 percentage points for the largest ones. The results reflect public sentiment during the time of data collection, which took place in the second half of January 2025.