
All opinion institutes failed to predict the Moderates' strong election results and how even the final election results would be.
AI-generated summary
Opinion polls measure voter support before elections, but may have difficulty capturing late voter transfers and support votes.
The Liberals' upswing was captured, and it was clear that the SD would stagnate or lose - but that the Moderates would succeed so well - it was missed by all opinion institutes.
The worst was Verian, where the Moderates got 15.8 percent - according to the election results on Monday morning, the party got 19.9 percent.
- It is the same pattern that was observed in the 2022 election. It is probably due to a strong final sprint where the focus was on the prime ministerial duels, says Per Oleskog Tryggvason, head of opinion at Indikator.
He emphasizes that on Monday it is a bit early to make financial statements - but at the same time he does not expect any major changes. This is what the list looks like - the number indicates the average deviation between the last opinion poll and the election result in percentage points:
1. Valu Svt's election survey: 0.6
2. Ipsos (DN): 0.8
3. Indicator (Echo): 0.9
4. TV4's Novus on election day and Verian/ (Svt): 1.1
6. Novus (Tv4) and Demoskop: 1.3
7. Sentio (National): 1.4
- The most prominent pattern is that the moderates were underestimated. It appears that as many as a third of voters made up their minds late and they went mostly to the right.
The opinion institutes had indicated that it would be even between the blocks - but not that it would be so even.
- The primary conclusion is that the measurement methods basically work well. Despite this precision, most institutes failed to predict how even the election would actually be, says Johan Martinsson, head of analysis at Demoskop.
This also applies to SVT's election poll Valu:
- They showed impressive precision with an average deviation of just 0.6 percentage points - but still gave a false picture of the block differences which looked like a clear victory for the opposition rather than the even result it turned out to be, says Johan Martinsson.
Jonas Hinnfors, professor of political science at the University of Gothenburg, also believes that the opinion institutes did their job.
- Valu missed the results of the Moderate Party and the Green Party, and there is reason to think about why. But you have to remember that opinion polls are not reality. It is a craft that has many sources of error. Overall, they succeeded well, says Jonas Hinnfors.
The most challenging thing for the opinion institutes has been the large, late voter movements that the rise of the Liberals created.
- We had them very low, along with many other institutes, just a few weeks ago. The upvote dynamic that follows is a challenge to capture. But we are very satisfied that we managed to capture that movement, says Per Oleskog Tryggvason.
Already today, the opinion institutes are starting their post-election surveys.
According to Johan Martinsson at Demoskop, the central questions are why the Moderates were so consistently underestimated, what proportion of the Liberals' votes were actual support votes and where the Sweden Democrats' expected voters went.

The preliminary voter turnout amounts to 80.3 percent but is expected to rise. The electoral authority explains that the figures for invalid votes and slow vote counting depend on procedures and logistics.

All over Sweden, voters are reacting to the even election results. Many express uncertainty about the formation of the government and the future, while others feel joy or relief depending on the political domicile.

The Left Party increased strongly in vulnerable areas during the 2026 parliamentary elections, especially in Rinkeby and Rosengård, while the Party of Nuance collapsed and its leader Mikail Yüksel resigned. The Social Democrats also increased in these areas, but not at the expense of the Left Party.
The election results in the parliamentary elections, the regional elections and the municipal elections in 2026 show that the future of the Social Democrats is challenged in Ånge and that the political majority is unclear, which could lead to S ending up in opposition for the first time.

Sunday's election in Sweden 2026 was a disappointment for many, with the former dominant party with Nazi roots backing down for the first time in a parliamentary election, while the Social Democrats made their worst election since universal suffrage was introduced. The article uses the fable of the hare and the tortoise – here transformed into a snail – as a metaphor for hubris and fall, and reflects on the fragility of democracy and the possibility of renewal.

After the 2026 election, Magdalena Andersson's government base leads with 176–173 seats over Ulf Kristersson's block, with a vote margin of only 0.45 percentage points or 28,600 votes. The late Wednesday votes could change the outcome, but history shows they rarely turn the tide of blocs. The Social Democrats are at around 28 percent, which would be their weakest election in over 100 years, while the Sweden Democrats are back compared to 2022 and the Moderates are the second largest party.