The whole term

Analysis

The whole term in one place: where the parties’ plans and votes part, what that would have cost, who votes with whom — and last the chamber and its members, straight from the voting data.

2,539All cases
20,288AI decisions
1,700Broken commitments
324Reversible decisions
10Government-side defeats

By policy area

Where do parties follow their plan — and where not?

Share of votes that followed the party's own plan, broken down by policy area. Parties sit at very different levels overall, so the columns differ sharply in tone — read each against that party’s own average in the footer row, and the “Biggest swings” table below for where a party varies most against itself.

Policy areaVSMPCLKDMSDn
Health & social services62%53%68%47%15%11%35%29%278
Environment & agriculture53%46%66%54%24%20%44%50%241
Justice63%53%58%38%22%25%36%44%223
Education62%63%62%54%21%24%47%40%217
Civil law & housing55%51%54%53%31%25%54%51%202
Business & energy69%62%61%54%38%26%54%55%196
Transport & infrastructure53%63%51%67%30%26%56%56%160
Constitution59%42%49%49%25%18%48%46%157
Migration & social insurance78%62%81%54%32%41%63%70%152
Foreign affairs64%61%69%50%20%22%51%68%131
Defence58%52%63%48%20%17%33%31%126
Culture39%42%59%37%10%8%51%36%119
Taxation70%61%64%64%34%35%42%54%117
Labour market49%68%58%61%28%25%48%58%116
Economy & finance69%66%60%53%41%39%63%55%104
Party average60%55%61%51%25%23%47%48%18011

The denominator is votes where the party took a side (Ja or Nej) — the same gate as the published divergence figure. Abstentions are not counted as misses: a p6 position can only resolve to Ja or Nej, so an area attracting many abstentions would otherwise look like one where parties abandon their plan. They are reported separately in the cell description instead. n = number of divisions in the area.

Colour is the cell’s own figure on a fixed scale: red = 0%, light green = 50%, green = 100%. The same figure gets the same colour everywhere, so columns can be compared directly. How an area sits against the party’s own average is in the cell’s tooltip and in the table below.

Largest swings per party

The area where each party deviates furthest above and below its own average.

Follows the plan mostFollows the plan least
VVänsterpartietMigration & social insurance +18 pp 78%Culture −21 pp 39%
SSocialdemokraternaLabour market +13 pp 68%Constitution −14 pp 42%
MPMiljöpartietMigration & social insurance +19 pp 81%Constitution −12 pp 49%
CCenterpartietTransport & infrastructure +16 pp 67%Culture −14 pp 37%
LLiberalernaEconomy & finance +16 pp 41%Culture −15 pp 10%
KDKristdemokraternaMigration & social insurance +18 pp 41%Culture −15 pp 8%
MModeraternaMigration & social insurance +15 pp 63%Defence −14 pp 33%
SDSverigedemokraternaMigration & social insurance +21 pp 70%Health & social services −19 pp 29%

Areas with fewer than 40 votes for that party are excluded from the ranking. The smallest cell in this run holds 72 votes, so nothing is actually filtered out.

Counterfactual outcome

Decisions that would have gone the other way

Divisions whose outcome would not have survived the parties voting their explicit commitments. Only parties whose own plan took a position outright (evidence tier "explicit commitment") are moved, and only where the party voted Ja or Nej. Every other party keeps its real seat counts, including its own dissenters. A party that abstained stays where it stood — an abstention is a floor tactic the plan was never asked to express.

324decisions that would have come out differently (of 2,539)
66where one party's commitments alone would have been enough
957divisions with at least one broken explicit commitment

Direction: 324 decisions go from Ja to Nej, 0 from Nej to Ja. A further 5 divisions end tied and are counted in neither direction — the Riksdag settles a tie by lot.

Which parties carry the reversal?

Votes movedReversed the decisionDecisive alone
LLiberalerna45818414
MModeraterna24517615
SDSverigedemokraterna22416824
KDKristdemokraterna3031425
VVänsterpartiet145512
CCenterpartiet109501
MPMiljöpartiet135441
SSocialdemokraterna81414

One decision can be carried by several parties at once, so the "reversed the decision" column sums to more than the number of reversals. The governing parties dominate for the same reason as everywhere else in this measurement: they are read against their own manifestos, not against Tidöavtalet.

The divisions (324)

2026-06-17Om återtagandet av eftersatt järnvägsunderhållTransport & infrastructureMSDJaNej
2026-06-17VägbelysningsnätetTransport & infrastructureMaloneJaNej
2026-06-09Riksrevisionens rapport om tillämpningen av det finanspolitiska ramverket 2025Economy & financeSSDJaNej
2026-06-09Subsidiaritetsprövning av förslag till förordning om den 28:e ordningens bolagsregelverkCivil law & housingSDaloneJaNej
2026-06-09SanktionerForeign affairsLSDJaNej
2026-06-09Ekonomiska konsekvenserEducationKDaloneJaNej
2026-06-03Straff och sanktionsavgifterEnvironment & agricultureSDaloneJaNej
2026-05-27JämställdhetForeign affairsKDLSJaNej
2026-05-20Ett starkt lokalsamhälle och en effektiv landsbygdspolitikBusiness & energyKDLJaNej
2026-05-20Offentlig och kommersiell serviceBusiness & energySDaloneJaNej
2026-05-20KompetensförsörjningHealth & social servicesCKDMPJaNej
2026-05-20Civilsamhällets organisationerBusiness & energyLaloneJaNej

Showing the 12 most recent of 324. See all in the case browser →

This is arithmetic, not a prediction. It shows what the chamber's numbers would have been had every explicit commitment been kept — not what would have happened politically. Had the parties voted differently, the committee would likely have written a different proposal, other parties would have reacted, and the negotiations would have looked different. Read the rows as questions about individual commitments, not as an alternative term of parliament.

Alignment

Who votes with whom?

Actual votes

VSMPCLKDMSD
V41%55%34%42%42%42%29%
S41%44%39%50%49%50%36%
MP55%44%39%37%37%37%27%
C34%39%39%50%50%50%42%
L42%50%37%50%100%100%79%
KD42%49%37%50%100%100%79%
M42%50%37%50%100%100%79%
SD29%36%27%42%79%79%79%

Share of divisions where two parties took the same position (Ja/Nej/Avstår), whole term.

If parties voted their own documents

The same measurement, taken on the position each party's own election manifesto and party programme point at instead of on its vote in the chamber. Cells sit in the same place in both tables — hover a cell for the real figure and the difference.

VSMPCLKDMSD
V86%94%72%67%66%42%53%
S86%86%73%75%77%54%64%
MP94%86%75%71%69%44%55%
C72%73%75%83%81%66%69%
L67%75%71%83%82%69%71%
KD66%77%69%81%82%71%78%
M42%54%44%66%69%71%79%
SD53%64%55%69%71%78%79%

Share of divisions where two parties' document-grounded positions coincided. Corner mark = at least 10 percentage points away from the real table. There is no absence here: every party has a plan for every division, so the denominator is all of them.

The coalition, or the party programme?

Every division pits the committee's proposal (Ja) against a reservation (Nej). Two rival explanations of how a party votes are set against each other: the coalition line — vote with the committee, knowing nothing about the party — and the party programme, which the AI agent reads without ever being told who authored the reservation. The gap between them is the measurement.

What explains the party's vote?

Coalition lineParty programmeWhat the programme addsn
KD100%23%−76.2 pp2536
L100%25%−74.2 pp2536
M100%47%−52.6 pp2536
SD79%44%−35.4 pp2536
C50%39%−10.6 pp2536
S50%48%−1.8 pp2536
V42%49%+6.9 pp2536
MP37%46%+8.8 pp2536

Coalition line = the share of votes hit by the rule "always vote Ja, with the committee". Party programme = how often the party's own plan points to the vote it actually cast. The difference is what the plan adds: zero means it explains nothing beyond coalition position.

M · KD · L — This is now measured without bloc documents: M, KD and L are read against their own manifestos and party programmes, not against the Tidö agreement. That the coalition line accounts for nearly 99% of their votes while their own plans often point elsewhere is therefore not an error in the measurement — it is the measurement. The difference is what the coalition deal carries, not the party programme.

When the coalition outweighs a party’s own documents

Divisions where the party’s own documents explicitly imply voting against the committee — yet it voted Ja anyway. Split by whether the party moved the reservation itself.

Own reservationAnother party's reservation
strictloosestrictloose
KD209/209 100%302/302 100%
L0/1 0%0/1 0%348/348 100%457/457 100%
M190/190 100%245/245 100%
SD0/89 0%0/92 0%199/239 83%224/270 83%
C0/142 0%0/158 0%89/194 46%102/218 47%
S0/202 0%0/219 0%63/118 53%79/146 54%
V0/208 0%0/214 0%135/241 56%144/256 56%
MP0/300 0%0/310 0%118/297 40%133/326 41%

Where a party moved the reservation itself, it never abandoned it — 0% for every party, at both thresholds. Procedure is decisive where procedure applies. The right-hand column is therefore the real test.

The divisions (1686)

The individual divisions are listed on each party's own page, alongside the rest of that party's figures.

What this does not show

A flag means a document-grounded reading of the party’s programme pointed the other way from its vote. It cannot by itself distinguish "coalition discipline overrode the programme" from "the AI agent misread the programme". That parties never abandon their own reservation (0% in every case below) suggests the measurement captures something real — but individual rows should be read as questions, not verdicts.

The chamber

Government-side defeats (10)

Computed directly from the Riksdag voting data — no AI involved. Source: The Swedish Parliament.

2025-05-22Organisationsform m.m.2024/25:UbU15 p.49Nej vs Ja
2025-04-24Kontantförbud vid handel med järn- och metallavfall2024/25:NU17 p.20Nej vs Ja
2025-04-02Beteslagstiftning och utevistelse2024/25:MJU10 p.12Ja vs Nej
2024-11-27Ett skuld- och kreditregister2024/25:CU5 p.4Nej vs Ja
2024-06-12Uppräkning av assistansersättning2023/24:SoU13 p.7Nej vs Ja
2024-05-15Översyn av sanktionerna mot terrorism2023/24:UU10 p.3Nej vs Ja
2023-03-30Myndigheternas roll vid tillståndsprocesser2022/23:NU13 p.5Nej vs Ja
2023-03-09Anläggande av våtmarker2022/23:MJU7 p.11Nej vs Ja
2023-03-08Ett långsiktigt hållbart transportsystem2022/23:TU4 p.4Ja vs Nej
2022-12-21Subsidiaritetsprövning av kommissionens förslag till förordning om inrättande av ett krisinstrument för den inre marknaden2022/23:NU10 p.1Nej vs Ja

Divisions where M, KD and L shared a line and the chamber decided the other way.