Tools chosen: polymarket_direct polymarket_related kalshi_related claude_news gdelt_news wikipedia code_execution

Sub-questions (Fermi decomposition)

Planner reasoning

This is a Polymarket question on the 2026 Mecklenburg-Vorpommern state election, where AfD has been leading state polls significantly. Key drivers are current polling margins, trends since the 2024/2025 federal shifts, and whether SPD (incumbent Schwesig) can close the gap. The market price is the primary anchor, supplemented by German polling news and cross-market signals.

# Current state AfD has led every published MV state poll throughout 2026, currently polling ~36-37% vs SPD's ~28-29% (a 7-9 point lead as of late August 2026), roughly three weeks before the September 20, 2026 election. The question resolves solely on seat count, not on government formation — a mainstream "firewall" coalition is expected to govern regardless, but that does not affect who wins the plurality of seats. # Timeline of key events - 2021-09-26 (confirmed): SPD won 2021 MV election with ~40%; AfD 2nd at ~17%; SPD-Left coalition formed under Schwesig. - 2026-02 (confirmed poll, Forsa): AfD 37%, SPD 23% — 14-point AfD lead. - 2026-05 (confirmed poll, Infratest dimap): AfD 36%, SPD 27% — 9-point lead. - 2026-06 (confirmed poll, INSA "late June"): SPD ticking up slightly per t-online reporting. - 2026-07-05/06 (confirmed): German Defence Minister Pistorius publicly raised alarm over possible far-right state government win in MV; protests against AfD reported in multiple cities. - 2026-07-05 (reported): AfD federal party conference held. - 2026-08-04/11 (confirmed poll, INSA): AfD 36%, SPD 29%, Linke 11%, CDU 10%. - 2026-08-11 (reported): Poll shows majority of Germans oppose the anti-AfD "firewall" nationally (Washington Examiner). - 2026-08-13 (reported): Left-wing parties pouring "record sums" into anti-AfD campaigning in eastern states (Zerohedge/Sott). - 2026-08-21/27 (confirmed poll, Forsa): AfD 37%, SPD 28%, Linke 11%, CDU 10%. - 2026-08-31 (reported): Merz warns of "angry citizen" sentiment at Saxony-Anhalt campaign stop (adjacent state, same election cycle context). - 2026-09-02 (reported): Atlantic Council notes stakes for AfD/Merz across three simultaneous September German state elections. - 2026-09-20 (scheduled, confirmed): Election day. # Event Will AfD win the most Landtag seats in the September 20, 2026 Mecklenburg-Vorpommern election? # Outcomes to forecast Yes (AfD wins most seats) / No (another party wins most seats) # Kalshi market anchor No Kalshi-direct price returned (tool found no matching MV/AfD tickers). Closest comparable Kalshi market data unrelated (German Chancellor/Weidel market at 15%, not comparable). **Primary cross-market anchor is Polymarket**, pricing this exact event: **88.5% YES**, down 2.5pts over 7 days but up 3.5pts over 30 days; range 68.5-93.5% over 91 days; volume $171K. # Sub-question answers 1. **Current AfD polling lead**: ~36-37% vs SPD ~28-29%, lead ~7-9 points (Forsa Aug 21-27, INSA Aug 4-11; dawum.de average: AfD 36.7% vs SPD 28.3%, gap 8.4pts). 2. **Trend over 6-12 months**: Lead narrowed from 14pts (Feb 2026 Forsa) to 9pts (May) to 7-9pts (Aug), with SPD ticking up modestly — narrowing but AfD remains clearly first, no reversal. 3. **Historical base rate**: No direct precedent cited in research; code_execution model estimates ~79-95% probability of a poll leader holding a 5-10pt lead through to election day at this distance, based on statistical error assumptions (not empirical German-specific base rate). 4. **2025 federal/2024 EU-local results in MV**: Not directly reported in raw research (no MV-specific 2025 federal or 2024 EU/local vote shares provided); Wikipedia confirms AfD was 2nd in 2021 MV state election at ~17%, later became 2nd-largest nationally after 2025 federal election, consistent with AfD strengthening trend but no MV-specific confirmation. 5. **Schwesig/SPD momentum**: SPD polling slightly up recently (t-online: "ein Prozentpunkt mehr" vs late June), but still trailing AfD by ~7-9pts with weeks remaining; no evidence of Brandenburg-2024-style late-surge dynamics materializing yet. 6. **Delay risk**: No indication of any scheduling risk; Wikipedia confirms election set for Sept 20, 2026, within legally permissible window (Aug 30–Nov 22, 2026); no rules/legal challenges reported. # Key facts (high-confidence, factual) 1. [dawum.de/claude_news] AfD polling average ~36.7%, SPD ~28.3%, as of Aug 2026 — 8.4pt lead. 2. [Wikipedia] 2021 election: SPD won with ~40%, AfD 2nd at ~17%. 3. [Wikipedia] Election legally scheduled/confirmed for Sept 20, 2026. 4. [t-online] CDU rules out coalition with AfD or Linke; firewall dynamics intact. 5. [Polymarket] Market prices this exact event at 88.5% YES currently. # Cross-market signals - Kalshi related: no direct match found. - Polymarket: 88.5% YES on identical event, high confidence, moderate volume ($171K), stable-to-rising trend over 30 days despite 7-day dip. - Sportsbook implied: none available. # Analyst opinions and speculation - Atlantic Council/media framing treats AfD win as base case ("what's at stake for AfD"), implying elite consensus expects AfD plurality (gdelt). - Zerohedge/Sott frame anti-AfD spending as defensive reaction to AfD "surge," reinforcing AfD-favored narrative. - code_execution model: pure poll-based probability 79-95% depending on lead-volatility assumptions; blending with hypothetical lower "market" price is speculative/illustrative, not evidence-based. # Directional lean per outcome - **Yes (AfD)**: Consistent, large (7-14pt), un-reversed lead across all 2026 polls; elite/media framing assumes AfD win; Polymarket at 88.5%. Opposing: lead has narrowed somewhat; MMP seat conversion and late tactical consolidation among anti-AfD voters could compress margin; no MV-specific historical base rate confirmed empirically. - **No (other party, likely SPD)**: SPD showing modest late uptick and Schwesig incumbency; anti-AfD "firewall" campaign spending record sums; 2021 precedent showed SPD outperforming pre-election polls significantly. Opposing: current gap (7-9pts) much larger than what SPD overcame in 2021 polling errors typically seen in Germany. # Gaps / unknowns - No MV-specific 2025 federal/2024 EU-local election results provided to validate polling. - No Kalshi-direct price for this specific ticker was returned — Polymarket used as sole cross-market anchor. - No empirical historical base-rate table for German state elections at this polling distance. # Calibration anchors - Polymarket YES price (best available cross-market anchor): 88.5%. - 2021 MV election: incumbent SPD outperformed some pre-election expectations, but AfD's current lead (~8pts) is far larger and more stable than any comparable pre-2021 SPD-AfD gap.

1. Both forecasters anchor heavily on Polymarket (88.5%) despite the brief explicitly noting Kalshi has no direct match and the code_execution model's own poll-based estimate range (79-95%) is wide and not empirically grounded in German-specific base rates — neither forecast explores the low end (79%) or explains why they land at the very top of that range rather than nearer the midpoint. 2. Neither forecaster addresses MMP seat-conversion mechanics in detail: Germany's mixed-member proportional system with overhang/leveling seats could produce a seat outcome diverging from vote-share plurality, especially if AfD's vote is concentrated in fewer constituencies or if a 5% threshold trips up smaller parties (Linke ~11%, CDU ~10% are close to relevant thresholds) — this could redistribute seats in non-obvious ways that pure vote-share polling doesn't capture. 3. Forecast 1 cites Brandenburg 2024 as a counter-example but dismisses it as "only ~5pts" without checking whether MV's electorate/demographics are comparable — this is a thin analogy given the brief explicitly says "No historical MV-specific base rate confirmed empirically" and "No direct precedent cited in research." 4. Both forecasts underweight the tail scenario of coordinated tactical voting/consolidation explicitly happening in real time: the brief notes "record sums" being poured into anti-AfD campaigning specifically in eastern states in the final weeks (Aug 13), which is a more direct and recent signal than the Brandenburg comparison, yet neither forecaster quantifies how much this could compress an 8-point gap in three weeks. 5. Neither forecaster flags the potential for a shy/differential turnout effect for AfD (historically AfD polls have sometimes under- or over-performed in different German elections) — the brief provides no polling-error track record for MV or comparable eastern states, and this omission from both forecasts leaves an unexamined source of variance that could push either direction. 6. Both converge to nearly the same number (0.87 vs 0.89) despite claiming somewhat different reasoning (haircut vs. no haircut), suggesting anchoring bias to Polymarket rather than independent evidence weighing — true independent analysis should show more dispersion given the acknowledged gaps in empirical base rates.

87% (71% confidence)

Yes: 87% No: 13%

AfD holds a stable, un-reversed 7-9 point polling lead across all 2026 polls three weeks out, and Polymarket prices this exact event at 88.5% YES with meaningful volume. The critique validly flags the narrowing trend (14pts to ~8pts), record anti-AfD spending, and the Brandenburg 2024 late-consolidation precedent — but that reversal overcame only a ~4-5pt gap, roughly half of AfD's current margin, and SPD's uptick so far is only ~1 point. MMP seat-conversion risk at an 8pt vote gap is minimal since proportional allocation dominates. I land at 0.87, a modest haircut from the market anchor to reflect the closing trend and unquantified late tactical-voting risk, but the critique does not justify moving toward the low end (~0.79) given the size and consistency of the lead.

Total pipeline time: 179.3s