Now you can explore in detail where those who decided came from.
Main Force: Awakening of Non-Voters
Election polls already before the elections essentially throughout the entire four-year term of Parliament recorded "strong mobilization of former voters of parties that did not succeed last time, toward ANO." Detailed analysis conducted using the ecological inference method now confirms these facts in full force: one million votes that were wasted last time turned into 700 thousand votes for Andrej Babiš this year.
SPD Got Vacuumed by ANO Again
That SPD was losing electoral potential showed up at the last minute, a few days before the elections the polls captured it, but only weakly. Reality was even more dramatic, however, due to the moratorium that legally prohibits publishing polls three days before elections, it didn't make it into public polls. This happens often: dynamics before elections are high, and polls have a chance to capture that voters are deciding until the last moments - in this case - between ANO and SPD. The closer the election date approached, the more voters leaned toward giving their vote to the ANO movement. A smaller part of SPD voters then went to the Motorists, who offered similarly protest rhetoric.
Stačilo!: Communists Without Social Democrats
Kantar's analysis before the elections showed that the Stačilo! grouping was planned to be supported by only a small part of SOCDEM voters; significantly more chose the ANO movement and a substantial part of them did not intend to come to the polls.
The data confirms this: KSČM voters from 2021 split: the larger half moved to ANO and only a minority decided for Stačilo!. Even greater losses are among social democrat voters – they did not support the SOCDEM connection with Stačilo! and mostly also voted for ANO. The result? Stačilo! ended just below 5% and remained outside Parliament. The left without social democracy simply does not have a sufficient voter base in Czechia.
Pirates: Resurrection Thanks to Young Non-Voters
Pirates have long been helped by the image of a party that has no problems with corruption and other scandals. Now their successful pre-election campaign also worked, many voters were impressed by the "green wave" thanks to the inclusion of Green Party representatives on candidate lists.
STAN: Silent Winner of Redistribution
The Mayors maintained voters from previous elections and further gains came from Spolu – especially from TOP 09 and KDU-ČSL, whose voters were looking for a centrist alternative with less ideological coloring and greater orientation toward local politics.
Spolu: Losses, Redistribution in the Democratic Block
The Spolu coalition lost voters on all fronts – they went to STAN, Pirates and also to the Motorists. Only a negligible part of them dropped out completely and didn't vote, and about 50 thousand Spolu voters from 2021 voted for ANO this time.
However, the context is crucial: Most Spolu voters remained in the "democratic block." They moved to STAN or Pirates, but did not go to ANO or protest parties.
Explanation? The "defeat Babiš" strategy stopped working as expected. In 2021, voters voted for Spolu strategically – they wanted to stop Andrej Babiš from governing again. This year, when defeating Babiš was not realistic and "Anti-Babiš" was not a campaign theme, they voted according to their actual preferences. Those who four years ago were deciding between Spolu and PirSTAN and chose Spolu because of the higher chance of defeating Babiš, voted more freely this year – STAN or Pirates.
Motorists: Perfect Mobilization Machine
Motorists to Themselves ran a campaign based on social networks and influencers. They gained:
- About 80 thousand new voters from non-voters
- Approximately 65 thousand from SPD
- Smaller parts from ANO (about 15 thousand) and from Spolu (about 10 thousand)
This diversification of vote sources secured their entry into Parliament with almost 7% and they became the key to a majority for a potential ANO-SPD-Motorists government.
Oath: A Bubble That Definitively Deflated
Róbert Šlachta's Oath consisted of so-called homeless voters. In 2021, it ended just below 5% with 251 thousand votes (4.68%). In 2025, it didn't even reach 3%. Oath had no solid voter core – what it collected broadly from everyone in 2021 (protest against the establishment), it broadly lost back this year. Part to ANO, part to Motorists, part back to non-voters. It was a purely protest vote that did not find a permanent anchor.
Paradox 2025: Fewer Wasted Votes, But the Same Problem
In 2021, more than a million votes were wasted – a record in Czech history. This year it was "only" around 300 thousand. Yet the question remains: Who represents the forgotten?
Wasted Votes: Comparison 2021 vs. 2025
2021 Elections: A record 1,069,359 votes were wasted (19.9% of all valid votes). It was the highest number of wasted votes in the history of independent Czechia.
Main "culprits" of waste in 2021:
- ČSSD: 250 thousand votes (4.65%)
- KSČM: 194 thousand votes (3.60%)
- Oath: 251 thousand votes (4.68%)
- Trikolóra, Svobodní, Soukromníci: 148 thousand votes (2.76%)
2025 Elections: Approximately 300-350 thousand votes were wasted (about 5-6% of all valid votes).
Main wasted votes:
- Stačilo!: about 220-250 thousand votes (4.7-4.9%)
- Oath: about 60-70 thousand votes (1.07%)
- Other small parties: the rest
What Changed in Four Years?
1. Consolidation of the Left (Which Failed)
In 2025, KSČM and SOCDEM merged into the Stačilo! grouping, but SOCDEM voters did not support this connection and mostly voted for ANO. Instead of saving the left, the merger documented its collapse.
2. SPD Got Into Parliament (But Weakened)
In 2021, SPD gained 9.56% and got comfortably into Parliament. In 2025, it gained 7.78% – still above the threshold, but significantly weakened due to the outflow of voters to ANO.
3. Motorists Replaced Oath
Oath in 2021 ended just below the threshold with 4.68%. In 2025, it fell to a mere 1.07%. But its place was taken by the Motorists (6.77%), who addressed a similar protest electorate, but with a better campaign.
Who Represents "Wasted Voters"?
In 2021, a million voters without representation posed considerable risk for the governing coalition parties. These were mainly voters with lower education and also from poorer regions. And they now showed that they do not wish for further governance by the Spolu coalition.
Historical context of wasted votes:
- 2006: 319 thousand (least in history)
- 2010: 985 thousand
- 2017: 317 thousand
- 2021: 1,069 thousand (record)
- 2025: about 300-350 thousand
How to Read Voter Shift Charts
🔶 The first view always shows where voters from the 2017 parliamentary elections moved.
🔶 The second then shows where parties collected voters in this year's elections.
🔶 Clicking on a colored rectangle on the sides of the chart displays the selected party or coalition.
How Voter Shifts Are Calculated
The method by which voter shift structures are revealed is called ecological inference. It is a fairly complex statistical method based on so-called Bayesian calculation, by which the most probable structure of voter shifts is determined according to the variance in the number of votes for individual parties. This happens in a large number of geographic units, e.g., in municipalities or polling stations.
In other words: Ecological inference is a statistical method that helps determine how voters moved between individual parties between elections – even though we don't have data about specific people (elections are of course secret).
It is based on the fact that we know election results for smaller territorial units, such as municipalities or polling stations. When we compare these results between the last elections (in our case 2021 and 2025), we can use a mathematical model to estimate what voter shifts are most probable. For example, that in municipalities where ODS previously had strong support and ANO newly succeeded, part of ODS voters could have moved - logically - precisely to ANO.
The model relies on so-called Bayesian calculation, which looks for the most probable explanation of observed results. In practice, various possible combinations of shifts are calculated and it is determined which of them best corresponds to real data from many municipalities.
The result is not one hundred percent certainty, but an estimate based on probability. The method is reliable mainly when we have a large amount of diverse data – that is, when results between municipalities differ a lot. If, on the other hand, all municipalities are similar or there is not enough information, the result may be less accurate.
In any case, it is the most used professional way to estimate voter shifts without asking people directly, and therefore it is used, for example, by political scientists, journalists, or election analysts around the world.
┃┃┃
Mahdalová & Škop – Data that show voter movements.