Will S, V, MP and C win at least 175 seats?
96.4 %- Data cutoff
- 9 September 2026
- Snapshot
- pvf-afd65e45
- Simulations
- 10,000
Polling data through .
Sweden · Parliamentary election 2026
Forecast terminal
A connected model from published opinion polls to vote shares, seats and possible paths to government. All probabilities describe simulation frequencies, never betting odds.
EVIDENCE BASEWhat the forecast is based on
40polls in the current model
8independent polling series
101,109interview responses in total*
32historical party outcomes
10,000seat simulations
*Respondents may appear in multiple polls; the sum is not a count of unique people.
Forecast questions
A precise answer needs a precise question
7 DEFINITIONS
All classified as MODEL
The questions are defined in advance so their outcomes can be checked after the election. Each probability is the share of simulations meeting the condition. It says nothing about which outcome you should favour.
Will S, V, MP and C win at least 175 seats?
96.4%
This measures seat strength. It does not mean C and V accept the same government.
Definition and uncertainty
Yes if S, V, MP and C together receive at least 175 of 349 seats in the Election Authority's final parliamentary result. The forecast is a model, not a market price or a promise.
Will the Tidö parties win at least 175 seats?
3.6%
L's threshold position is the largest individual risk to the declared seat grouping.
Definition and uncertainty
Yes if M, SD, KD and L together receive at least 175 seats in the final result. The forecast is a model, not a market price or a promise.
Will the Liberals clear the parliamentary threshold?
6%
The threshold makes small movements around four percent very important for seat allocation.
Definition and uncertainty
Yes if the Liberals receive at least 4.00 percent of valid votes nationally. The forecast is a model, not a market price or a promise.
Will the Christian Democrats clear the parliamentary threshold?
97.9%
The model weighs the polling centre against historical forecast errors and variation between pollsters.
Definition and uncertainty
Yes if the Christian Democrats receive at least 4.00 percent of valid votes nationally. The forecast is a model, not a market price or a promise.
Will the Social Democrats be the largest party?
98.3%
Being the largest party affects negotiations but does not automatically appoint the prime minister.
Definition and uncertainty
Yes if the Social Democrats receive the highest national vote share among the modelled parliamentary parties. The forecast is a model, not a market price or a promise.
Will SD win more seats than M?
75.3%
This affects the balance of power within the right-wing grouping but does not automatically determine the prime minister.
Definition and uncertainty
Yes if the Sweden Democrats receive more parliamentary seats than the Moderates in the final result. The forecast is a model, not a market price or a promise.
Will exactly seven parties enter Parliament?
91.1%
In the central forecast, L loses its seats while the other seven parties win representation.
Definition and uncertainty
Yes if exactly seven of today's eight parliamentary parties win at least one seat. The forecast is a model, not a market price or a promise.
Party forecast · votes and seats
Central estimate, range and threshold risk
10,000 OUTCOMES
80% model intervals
Vote intervals reflect variation between polls and historical forecast errors. Each simulated vote share is then passed through the seat engine, so seats can change abruptly around the electoral threshold.
Social Democrats
29 %26.7–31.3 %
107median 106
98–114 seats
>99.9 %model probability
Sweden Democrats
19.4 %16.5–22.3 %
71median 71
60–82 seats
>99.9 %model probability
Moderates
17.4 %15.7–19.1 %
64median 64
58–70 seats
>99.9 %model probability
Left Party
8.1 %7.1–9.1 %
30median 30
26–33 seats
>99.9 %model probability
Centre Party
7.3 %6.2–8.5 %
27median 27
23–31 seats
99.6 %model probability
Christian Democrats
6.3 %5–7.6 %
23median 23
18–28 seats
97.9 %model probability
Green Party
7.5 %5.8–9.1 %
27median 27
21–34 seats
98.5 %model probability
Liberals
2.8 %1.8–3.7 %
0median 0
0–0 seats
6 %model probability
Forecast sensitivity
What is driving the outcome now?
DIAGNOSTICS
Derived from the same snapshot
A guide to the forecast: the party closest to the threshold, the likely largest party, the seat grouping closest to 175 and the widest seat interval.
6 % above the parliamentary threshold
Small movements around four percent can shift many seats between the blocs. This is the model's most sensitive threshold at present.
98.3 % simulation frequency
Social Democrats is largest in this share of simulations. That strengthens a negotiating position but does not appoint a prime minister.
S + V + MP
The central scenario is 11 seats from the majority threshold. Political compatibility is treated separately as sourced context.
Sweden Democrats
The widest 80% interval shows where seat outcomes vary most across simulations. Outcomes beyond the interval remain possible.
Government formation · model + dated context
Seats alone are not enough
MODEL / DECLARED / CONTEXT
Separate evidence categories
The coalition rows show the arithmetic. Below them are party leaders and public positions on government formation, with dates and source links. We do not assign probabilities to ministerial posts without a testable candidate set and a separate model.
S + V + MP + C reaches an outright majority in
96.4 %of simulations · not prime minister oddsM + SD + KD + L reaches an outright majority in
3.6 %of simulations · not prime minister odds191/ 349 seats in the central scenario96.4 % simulation frequency of at least 175
The seats may suffice, but C rejects V in government. A majority is therefore not a completed coalition.
An outright majority describes seat arithmetic. Forming a government also requires party conditions to be compatible or tolerated.164/ 349 seats in the central scenario7.9 % simulation frequency of at least 175
Seats for three red-green parties. The model does not assume other parties will tolerate the government.
An outright majority describes seat arithmetic. Forming a government also requires party conditions to be compatible or tolerated.161/ 349 seats in the central scenario3.7 % simulation frequency of at least 175
An Andersson grouping without V in government; it normally requires external support or tolerance.
An outright majority describes seat arithmetic. Forming a government also requires party conditions to be compatible or tolerated.157/ 349 seats in the central scenario2.1 % simulation frequency of at least 175
C has suggested this centrist solution, but the parties have no joint government agreement.
An outright majority describes seat arithmetic. Forming a government also requires party conditions to be compatible or tolerated.158/ 349 seats in the central scenario3.6 % simulation frequency of at least 175
The four Tidö parties' seat grouping. This is a seat forecast, not a settled government decision.
An outright majority describes seat arithmetic. Forming a government also requires party conditions to be compatible or tolerated.158/ 349 seats in the central scenario3 % simulation frequency of at least 175
Shows the seat position without L; it is not a separately declared government alternative.
An outright majority describes seat arithmetic. Forming a government also requires party conditions to be compatible or tolerated.DECLARED · CONTEXT
Party leaders and stated positions on government
Dated public statements, checked 6 Sept 2026. These are sourced context, not model odds.
Magdalena AnderssonParty chair and prime minister candidate
Leadership source · ↗- DECLAREDS seeks a mandate for an Andersson-led government
Magdalena Andersson says she seeks voters' mandate to become prime minister and lead a government pursuing broad cooperation.
The statement identifies a prime minister candidate but does not fix which parties would join or tolerate the government after the election.Social Democrats · effective from ↗
Ulf KristerssonParty leader and prime minister candidate
Leadership source · ↗- DECLAREDM campaigns for Kristersson and a four-party government
Ulf Kristersson has declared himself the Tidö grouping's prime minister candidate and said that M, SD, KD and L will join the government if the grouping wins.
This is a publicly declared government alternative. The model can test its outright seat majority, but cannot determine whether all post-election negotiations succeed.SVT Nyheter · effective from ↗
Jimmie ÅkessonParty chair
Leadership source · ↗- DECLAREDSD is to enter government if the Tidö parties win
Åkesson and Kristersson have presented an alternative including all four Tidö parties, with SD taking government responsibility.
The statement gives SD ministerial posts but does not specify individuals or exact portfolios. Those details are not modelled.SVT Nyheter · effective from ↗
Nooshi DadgostarParty chair
Leadership source · ↗- DECLAREDV demands cabinet participation if its votes are needed
The congress-adopted election platform says V will not support or permit a government it does not join if V's votes are needed to form it.
S, V, MP and C may reach 175 seats without their stated conditions being compatible. An outright majority is therefore not a completed government agreement.Left Party · effective from ↗
Elisabeth Thand RingqvistParty leader
Leadership source · ↗- DECLAREDC names Andersson as its most likely partner
Elisabeth Thand Ringqvist has identified Magdalena Andersson as her most likely prime minister partner. C demands lower emissions, more jobs, lower taxes and investment in rural areas.
C may be decisive in tolerating a prime minister even outside government. Its vote against a nominee can therefore matter more than whether a four-party grouping reaches 175 seats.SVT Nyheter · effective from ↗ - DECLAREDC wants to unite S, C and KD in the centre; MP has also been mentioned
C's proposed centrist alternative includes S, C and KD, with a possible opening for MP. MP has rejected the alternative.
This is C's proposal. There is no joint agreement, and KD has since refused to permit an Andersson-led government.Aftonbladet · effective from ↗ - DECLAREDC wants a centrist government without V or SD influence
The Centre Party campaigns for a centrist alternative without influence from the Left Party or Sweden Democrats.
C's conditions must be considered alongside V's and MP's demands for government participation. A combined seat count does not imply an agreement.Centre Party · effective from ↗
Ebba BuschParty leader
Leadership source · ↗- DECLAREDKD campaigns for continued Tidö cooperation
Ebba Busch has said KD supports Tidö cooperation and has no red lines about which Tidö parties may join the government.
KD is part of Kristersson's declared seat grouping, but specific ministerial posts are explicitly not assigned before the election.SVT Nyheter · effective from ↗ - DECLAREDKD rejects Andersson, even if it leads to an extra election
Ebba Busch has said KD is prepared to vote against Magdalena Andersson as prime minister all the way to an extra election.
According to this statement, C's preferred centrist government lacks KD's support. It is a dated political condition, not a model forecast.SVT Nyheter · effective from ↗
Amanda Lind & Daniel HelldénCo-spokespersons
Leadership source · ↗- DECLAREDMP demands a place in a red-green government
Amanda Lind has said MP must join a future red-green government and will vote against governments it does not join.
MP's seats therefore cannot automatically be counted as tolerance for an S-led government excluding MP.TV4 Nyheterna · effective from ↗ - DECLAREDMP rejects C's S + C + KD alternative
Amanda Lind has rejected the C leader's preferred centrist government of S, C and KD, citing differences in climate policy.
C's preferred centrist government therefore lacks a joint declaration from the parties it seeks to unite.Aftonbladet · effective from ↗
Simona MohamssonParty chair
Leadership source · ↗- DECLAREDL accepts SD in a four-party government
The Liberals' extraordinary congress confirmed the new position: the party is ready to admit SD to government if the Tidö parties win.
This removes the previous red line against SD ministers, but L must also clear the parliamentary threshold to win its own seats.SVT Nyheter · effective from ↗
The prime minister appoints ministers after the vote
Once Parliament approves the Speaker's prime minister nominee, the elected prime minister appoints the other ministers. Seats confer negotiating strength but do not automatically determine who receives a ministerial portfolio. Politicalverse therefore assigns no probabilities to individual future ministers.
Constitutional context from Swedish Parliament ↗The register describes constitutional rules and dated public statements. It does not model prime ministers, governing parties or ministerial posts and contains no odds for individuals.
Forecast history
New data points, visible changes
FROM 22 APRIL 2026
Through the data cutoff 9 September 2026
The history uses only information published by each date. It helps distinguish a stable forecast from a movement driven by a single new poll.
S + V + MP + CM + SD + KD + L 175 seats
First reported point22 Apr 2026S + V + MP + C 191M + SD + KD + L 158
Latest data point9 Sept 2026S + V + MP + C 191M + SD + KD + L 158
Each point is recalculated using only information published by that date. The line shows the model's central seat estimate, not observed election results.
Historical validation
We show where the model was wrong
4 COMPARABLE ELECTIONS
32 party outcomes
For each historical election, the data cutoff is the same number of days before election day as today's forecast. The forecast is then compared with the official result, party by party.
2010cutoff 15 Sept 2010
CALIBRATION
65 polls8 pollsters
0.75pp
S -1.39forecast minus result
2014cutoff 10 Sept 2014
CALIBRATION
57 polls9 pollsters
1.32pp
MP +2.71forecast minus result
2018cutoff 5 Sept 2018
CALIBRATION
74 polls9 pollsters
1.47pp
S -3.81forecast minus result
2022cutoff 7 Sept 2022
LOCKED HOLDOUT
99 polls8 pollsters
0.83pp
V +1.4forecast minus result
Comparable outcomes from 4 elections.
Percentage points across all reported party outcomes.
Historical coverage in leave-one-election-out validation. An 80% interval is not a guarantee.
How to read the validation: The frozen design is assessed against 2010–2018, with 2022 reserved as a locked holdout. The repository reproduces the fixed model, not an exhaustive parameter search. Four elections are a small sample; historical errors may understate new political shifts.
Method and data integrity
From a raw row to a reproducible snapshot
OFFICIAL data, POLLs, MODEL outputs and political CONTEXT are kept separate. This section shows the flow, version identifiers, cross-checks and limitations accompanying publication.
Published polls
The current window includes 40 polls from 8 polling series. A poll becomes available on its publication date.
Time and pollster
Newer fieldwork carries more weight, with a 28-day half-life. One pollster receives at most 25 % of the realised weight.
Historical forecast error
4 comparable elections inform the parties' joint uncertainty. 2002 and 2006 remain in the official history but are excluded from the eight-party test.
Seats, every time
10,000 possible vote-share outcomes pass through the same tested seat engine, with 29 constituencies and adjustment seats.
INTEGRITY RECORD
Every forecast can be reproduced.
The snapshot, source dataset, code version and checks are shown together. If an update fails validation, the last accepted forecast stays published.
Download snapshot (JSON) →- Classification
- MODEL · BETA
- Snapshot
pvf-afd65e45- Model version
1.0.0-beta.1- Data cutoff
- 9 Sept 2026
- Generated
- 9 Sept 2026 18:45
- Raw data
- 2,652 rows · from 1944
- SHA-256
2c5e7632f29c5aef732d2b68771cd02bf0fa47b3070cb0b5af6a850dea69fdef- Source revision
da8f8ea2857dff0f0016b9edd353616bc09512c9- Source adapter
pv-swedishpolls-1.1.0- Ties requiring a draw
- 138 simulations · 1.4 %
POLL · CC0SwedishPollsDataset and documentation ↗PRIMARY SOURCE CHECK · 19 Aug 2026NovusOpen the published poll ↗PRIMARY SOURCE CHECK · 17 Aug 2026SVT/VerianOpen the published poll ↗PRIMARY SOURCE CHECK · 4 Jun 2026SCBOpen the published poll ↗PRIMARY SOURCE CHECK · 2 Sept 2026NovusOpen the published poll ↗PRIMARY SOURCE CHECK · 3 Sept 2026SVT/VerianOpen the published poll ↗PRIMARY SOURCE CHECK · 3 Sept 2026DemoskopOpen the published poll ↗PRIMARY SOURCE CHECK · 4 Sept 2026IndikatorOpen the published poll ↗
Source correction: Novus, fieldwork 24 Aug 2026–30 Aug 2026. Publication date is recorded as 2 Sept 2026 according to the primary source ↗, instead of the aggregate file's 1 Sept 2026. The original file is preserved.
Limitations accompanying every number- Four elections have fully comparable eight-party polls. The frozen design is assessed on 2010–2018; 2022 is a locked holdout. OTHER is a combined residual category assumed to contain no eligible new party. Historical errors offer no guarantee for 2026, and seats are not the same as governing power.
- Other is an uncertainty category that keeps vote shares exhaustive, but receives no seats in v1. The entry of a new named party is therefore not modelled.
- The interview total does not count unique people; the same respondent may appear in multiple polls.
- Polls describe opinion during fieldwork. Campaign events after the data cutoff are not included in the forecast.
- The model calculates seats for government alternatives. Party conditions are presented as dated, sourced context without invented probabilities.
- Exact comparison-figure ties are resolved by a seeded, reproducible draw in the simulation. The record shows the frequency of elections requiring such a draw.