Scan: Will AfD win the most seats in 2 German state elections in September 2026?

1. Planner Fable 5.1 · Fermi decomposition + tool plan

Tools chosen: polymarket_direct polymarket_related kalshi_related claude_news article_search gdelt_news wikipedia

Sub-questions (Fermi decomposition)

What is the current Polymarket price for this market and its sibling markets (AfD wins most seats in 0, 1, or 3 September 2026 state elections)?
What do the latest Saxony-Anhalt state polls show for AfD versus CDU, and how large is AfD's lead?
What do the latest Mecklenburg-Vorpommern state polls show for AfD versus SPD/CDU, and how stable is AfD's lead?
What do the latest Berlin Abgeordnetenhaus polls show — is AfD anywhere near first place versus CDU, Greens, SPD, and Linke?
What are the confirmed election dates for the three elections, and is there any risk of postponement past Dec 31, 2026?
Have national AfD polling trends since 2025 (e.g., AfD vs CDU/CSU nationally) shifted in a way that would change the eastern state leads or make Berlin competitive?
How did AfD perform in the most recent state elections in these three states (2021 results) and how well did polls predict the winner?

Planner reasoning

This is a Polymarket question about AfD winning the most seats in exactly 2 of the 3 September 2026 state elections (Saxony-Anhalt, Mecklenburg-Vorpommern, Berlin). AfD leads polls strongly in Saxony-Anhalt and Mecklenburg-Vorpommern but is far behind in Berlin (CDU/Greens/SPD/Linke lead), so the key uncertainty is whether AfD holds first place in both eastern states and whether Berlin remains out of reach. Need current market price, latest state-level polling, and cross-market signals for the sibling outcomes (0, 1, 3).

2. Research Parallel tool calls

tool	status	time	items	summary
polymarket_direct	OK	0.4s	1	## This Market's Polymarket Data Will AfD win the most seats in 2 German state elections in September 2026? - Current price (probability): 72.50% - 7-day price change: -9.00% - 30-day price change: -3.50% - Total volume: $50,545 (USD notional) - Price range: 46.00% - 89.00% - Data points: 64 da
polymarket_related	OK	6.1s	0	Scanned 100 active Polymarket markets, kept 0 matches. keyword 'AfD state elections': 0 markets \| keyword 'Sachsen-Anhalt': 0 markets \| keyword 'Mecklenburg-Vorpommern': 0 markets \| keyword 'Berlin election': 0 markets \| keyword 'AfD': 0 markets
kalshi_related	OK	5.9s	1	1 related markets / summaries. keyword 'AfD': no matches \| keyword 'Germany state election': ok \| keyword 'Saxony-Anhalt': no matches
claude_news	OK	28.7s	12	## Key Findings Confirmed election dates: - Sachsen-Anhalt: 6 September 2026 (already held) The election to the ninth Landtag of Saxony-Anhalt took place on 6 September 2026. - Mecklenburg-Vorpommern: 20 September 2026 (upcoming) The ninth state election in Mecklenburg-Vorpommern wil
article_search	OK	0.8s	9	Fetched 9 unique articles across 4 queries (lookback=60d). 'AfD Sachsen-Anhalt Umfrage Landtagswahl 2026': 10 articles \| 'AfD Mecklenburg-Vorpommern poll 2026 state election': 10 articles \| 'Berlin Abgeordnetenhaus Wahl 2026 Umfrage CDU AfD': 10 articles \| 'AfD polls east Germany state elections Sep
gdelt_news	OK	166.4s	20	GDELT: 20 articles across 3 queries (lookback=45d). 'AfD Landtagswahl Sachsen-Anhalt 2026': 10 hits \| 'AfD Landtagswahl Mecklenburg-Vorpommern 2026': 10 hits \| 'Abgeordnetenhauswahl Berlin 2026': error GDELT rate-limited after retries (429)
wikipedia	OK	0.2s	4	Fetched 4 Wikipedia entries (1 missing pages).

3. Evidence Brief Sonnet 5 · 6715 chars

# Current state Sachsen-Anhalt already voted (6 Sept 2026) and AfD won decisively (43.8%, first place by a landslide) — that outcome is confirmed/official. Mecklenburg-Vorpommern and Berlin vote 20 Sept 2026: AfD is the clear polling leader in MV (36-37% vs SPD 28-29%), while in Berlin AfD is polling 2nd/3rd tier (18-22%), behind Die Linke, and roughly tied with CDU — not leading. Market resolves YES if AfD wins/ties for most seats in ≥2 of the three elections. # Timeline of key events - 2026-07/08: National/regional press notes record spending by left parties to stop AfD in eastern states; AfD polling strongly in SA and MV (reported, gdelt/zerohedge/sott). - 2026-08-11: Poll reported majority of Germans oppose the "firewall" against AfD cooperation (reported, Washington Examiner). - 2026-08-24/25: Infratest-Dimap poll: AfD 42% in Saxony-Anhalt (reported, CBC/Infratest-Dimap). - 2026-08-31: Merz booed/heckled at SA campaign stop; tension around AfD momentum (confirmed, AA/Yahoo). - 2026-09-02: Atlantic Council overview framing "three September elections" as pivotal test for AfD/Merz (reported). - 2026-09-06: Saxony-Anhalt election held — AfD wins with 43.8-44.4% (exit poll/final tallies vary slightly by source), CDU collapses to 17.2-18.3%; AfD falls a few seats short of absolute majority (39/83) (confirmed, Wikipedia/CNN/CBS/CNBC). - 2026-09-02 (Forsa/Ostsee-Zeitung, published): MV poll shows AfD steady at 37%, SPD up to 28% (reported). - Earlier MV poll (pre-Feb 2026 baseline vs later): AfD 36%, SPD 29%, gap narrowing ~5pts over the campaign (reported). - 2026-08-22 (Forsa) and 2026-08-31 (Civey): Berlin polls show Die Linke leading (20-22%), CDU and AfD roughly tied for 2nd/3rd (18-19%), Greens 4th (15-16%), SPD 5th (12-13%) (reported). - 2026-09-20: MV and Berlin elections scheduled (not yet held as of research cutoff). # Event Will AfD win or tie for the most seats in at least 2 of 3 September 2026 German state elections (Sachsen-Anhalt, Mecklenburg-Vorpommern, Berlin)? # Outcomes to forecast - Yes (AfD wins/ties most seats in ≥2 of 3 elections) - No (AfD wins in 0 or 1 of 3) # Kalshi market anchor No Kalshi-direct price returned; Kalshi_related search found no matching AfD/state-election ticker (only tangential Chancellor/redistricting markets). **Primary anchor is Polymarket**: sibling market currently priced at 72.5% YES, down from a 30-day high of 89% and a low of 46%; 7-day trend -9%, 30-day trend -3.5%; volume ~$50.5k over 64 days of data — moderate liquidity, some recent softening despite SA already being a confirmed AfD win. # Sub-question answers 1. **Polymarket price/siblings** — This market itself (2-of-3) trades at 72.5% YES (down 9% in 7 days). No sibling (0/1/3-election) markets were found in the polymarket_related scan. 2. **Saxony-Anhalt polls vs result** — AfD polled ~40-42% pre-election (Infratest-Dimap 42%, late Aug), consistent with the actual landslide result of 43.8% vs CDU 17.2%. AfD's win here is now CONFIRMED official result, not a forecast. 3. **MV polls** — Forsa poll (published 2026-09-02) has AfD steady at 37%, SPD gaining to 28% (up 5pts from Feb baseline); gap narrowing but AfD retains a clear ~8-9pt lead just before the Sept 20 vote. 4. **Berlin polls** — AfD is NOT leading; Forsa (~Aug 22) has Die Linke first at 22%, CDU/AfD tied 2nd at ~18-19%; Civey (Aug 31) shows Linke 20%, CDU 19%, AfD 18% — AfD in a competitive but sub-first-place position, unlikely to win outright. 5. **Election dates/postponement risk** — Sachsen-Anhalt already held Sept 6, 2026 (confirmed). MV and Berlin both scheduled Sept 20, 2026 (confirmed, Wikipedia). No reporting suggests postponement risk past Dec 31, 2026. 6. **National AfD trend shift** — Research shows no signal Berlin has become AfD-competitive; national narrative (record anti-AfD spending, firewall debate, polling on "firewall" unpopularity) suggests AfD strength concentrated in eastern states, not translating to Berlin's more urban/left-leaning electorate. 7. **2021 comparison / poll accuracy** — Not directly detailed in research, but SA 2026 pre-election polling (Infratest-Dimap ~42%) closely tracked the actual 43.8% result, suggesting good recent polling accuracy in this environment (CBC, Wikipedia). # Key facts (high-confidence, factual) 1. [Wikipedia] Sachsen-Anhalt election held Sept 6, 2026; AfD won 43.8% (or 44.4% per exit poll/CNBC), CDU 17.2-18.3%; AfD narrowly missed absolute majority (39/83 seats). 2. [Wikipedia] MV and Berlin elections both scheduled for Sept 20, 2026. 3. [Forsa/t-online, 2026-09-02] MV poll: AfD 37%, SPD 28%. 4. [Forsa/Civey via t-online] Berlin polls (Aug 22 & 31): Die Linke leads 20-22%, CDU/AfD tied ~18-19%. 5. [CBS/CNN] AfD won all but 3 of 41 SA single-member constituencies (won by The Left). # Cross-market signals - Kalshi related: no direct or closely comparable AfD state-election market found; only tangential German Chancellor/Weidel market (15% YES) exists. - Polymarket: this event's own contract at 72.5%, off recent highs (89%) — likely reflects narrowing MV/Berlin uncertainty priced in even after SA win locked in. - Sportsbook implied: none available. # Analyst opinions and speculation - Atlantic Council, CNN, CBS frame the three elections as a broader test of AfD's rise and the "firewall" strategy; commentators (Washington Post) see SA result as reflecting fear of change nationally. - Some outlets (zerohedge, countercurrents) frame AfD's East German wins as near-inevitable; treat as partisan/speculative framing, not structural fact. # Directional lean per outcome - **Yes**: SA win is confirmed (1 of 2 needed). MV polling shows stable ~9pt AfD lead close to election day — strong, but not certain, second win. Combined probability of ≥2 wins appears high. - **No**: Requires AfD losing MV (against consistent polling lead) since Berlin win is highly unlikely. Only plausible route to "No" is an MV upset (SPD momentum, narrowing 5pts) or turnout/tactical-voting effects unaccounted for in polls (as seen with SPD/Greens in SA beating low polls via tactical voting). # Gaps / unknowns - No Kalshi-direct YES price returned for this specific ticker; anchor relies on Polymarket only. - No polling data found analyzing undecided/tactical voting dynamics specific to MV close to Sept 20. - Sub-question 7 (2021 results comparison) not directly answered in research. # Calibration anchors - Polymarket YES price: 72.5% (down from 89% high, up from 46% low) — treat as consensus anchor. - SA precedent: pre-election polls (Infratest-Dimap 42%) closely matched actual result (43.8%), suggesting good polling reliability in this cycle, supporting confidence in MV's current AfD lead holding.

4. Forecasters Fable 5.1 + GPT-5.6 Sol · in parallel

claude-fable-5-1

56% conf

Yes	79%
No	21%

Sachsen-Anhalt is a confirmed AfD landslide (43.8% vs CDU 17.2%), so the question reduces almost entirely to Mecklenburg-Vorpommern on 20 Sept, since Berlin polling (AfD 18-19%, behind Die Linke 20-22%) makes a Berlin win very unlikely. The latest Forsa MV poll (2 Sept) has AfD 37% vs SPD 28%, a ~9-point lead that would normally be very safe three weeks out, and SA polling tracked the result closely this cycle. The main risk is a 2021-style late SPD/Schwesig consolidation (SPD massively outperformed polls then) plus tactical anti-AfD voting, which the narrowing 5-point trend hints at. Polymarket's 72.5% (down from 89%) is the anchor; I lean modestly above it given the size of the AfD lead and confirmed SA win, but keep meaningful weight on an MV upset.

gpt-5.6-sol

76% conf

Yes	79%
No	21%

Saxony-Anhalt has already supplied one confirmed AfD win, so the event now primarily depends on Mecklenburg-Vorpommern. AfD's consistent 36-37% polling there, roughly 8-9 points ahead of the SPD, points toward Yes, especially because recent Saxony-Anhalt polling closely matched the result. Berlin polling points toward No for that state, but it is mainly a backup route rather than necessary for resolution. With no Kalshi-direct price, I anchor on Polymarket's 72.5% Yes and move modestly upward because the confirmed first win and strong MV lead make an MV upset the principal remaining failure mode.

5. Devil's Advocate Sonnet 5 · red-team critique

Both forecasts converge on identical 79/21 splits despite different confidence levels, but neither explicitly quantifies the specific 2021 precedent risk they cite—the brief notes "SPD massively outperformed polls" in 2021 MV, yet no forecaster estimates how large that historical polling miss was or whether an ~9pt gap is large enough to survive a similar-magnitude swing (this is the single biggest identified risk and deserves explicit numeric treatment, not just narrative mention).

Neither forecaster addresses seat-allocation mechanics distinct from vote-share polling: German mixed-member systems with direct-mandate seats (as SA showed, "AfD won all but 3 of 41 single-member constituencies") mean a close SPD/AfD vote-share race in MV could still produce a clear AfD seat lead even if the poll gap narrows further—this actually argues for higher confidence than pure vote-share polling suggests, an upside case both missed.

Both forecasters treat Berlin as essentially a non-factor ("backup route"), but neither examines the tail scenario where AfD ties or narrowly edges CDU into 2nd in Berlin AND MV somehow fails—the compound "double failure" path is dismissed too quickly without checking correlation between MV and Berlin AfD momentum (a systemic national swing could move both simultaneously in either direction).

Both anchor on Polymarket 72.5% and land at exactly 79%, but neither justifies why the same ~6.5pt premium over the anchor is appropriate given Forecast 1 expresses more hedging language ("keep meaningful weight on an MV upset") while Forecast 2 sounds more committed ("principal remaining failure mode")—the inconsistency between stated confidence (0.56 vs 0.76) and identical output probabilities is unexplained and suggests one of them mis-stated their confidence or the rationale doesn't match the number.

Neither forecast addresses the Polymarket trend signal itself: the market is down 9% over 7 days and 3.5% over 30 days *despite* SA's win being locked in during that window, which should be read as the market pricing in something bearish for MV specifically (e.g. late-breaking SPD consolidation signals) that isn't fully reflected in the static 37/28 Forsa snapshot—this trend deserves more weight as a leading indicator than either gives it.

Both rationales lean on "SA polling tracked the result closely" as reassurance for MV polling reliability, but this is a single data point from a different state with a different (larger, more landslide-scale) AfD lead; extrapolating polling accuracy from a 42%-vs-43.8% miss to confidence in a 37%-vs-28% closer race is a reference-class stretch neither forecaster flags as uncertain.

6. Aggregator Fable 5.1 · final calibrated forecast

77% (66% confidence)

Yes: 77% No: 23%

Sachsen-Anhalt is a confirmed AfD win, so resolution hinges on Mecklenburg-Vorpommern on 20 Sept, where AfD holds a stable ~9pt lead (37 vs 28) in the latest Forsa poll; Berlin (AfD 18-19% behind Linke 20-22%) is only a low-probability backup route. The critique correctly notes the 2021 MV precedent deserves quantification: SPD outperformed polls by roughly 7-9 points then, which would nearly erase a 9pt gap — but that surge rode a coinciding federal election and Scholz coattails absent this cycle, and SA polling this month tracked the result closely. The Polymarket slide to 72.5% despite SA being locked in is a mild bearish signal about MV consolidation that both forecasters underweighted, so I shade slightly below their 79% while staying above the market given the size and stability of the AfD lead.

Pipeline Timing

Total pipeline time: 269.3s

Per-tool research timings shown in the Research section above.