Tools chosen: polymarket_direct wikipedia claude_news gdelt_news polymarket_related kalshi_related code_execution

Sub-questions (Fermi decomposition)

Planner reasoning

This market resolves on which party GAINS the most seats relative to its pre-election total, not which party wins the most seats — a crucial distinction, since United Russia already holds ~324/450 Duma seats and actually lost seats in 2021 while New People gained 13 from zero. Key research: current Polymarket price anchor, the exact pre-election seat baselines for each party, historical seat-change patterns, and whether the September 2026 election is on schedule and whether ER is expected to expand or shed seats (including annexed-territory seats and Kremlin party-management strategy).

# Current state Russia's 2026 State Duma election is confirmed for 18–20 September 2026 (Putin decree, per Interfax) — voting has not yet occurred. United Russia (ER) currently holds a Duma seat count reported variously as 314–324 of 450 (churn from byelections/defections since 2021), with 2021's official baseline being 324 seats. The market resolves on which party GAINS the most seats vs. its pre-election total — not simply which party wins the most total seats — a distinction sources conflate. # Timeline of key events - 2021-09-19: ER wins 2021 election with 324/450 seats (down from 343 in 2016); CPRF 57, SR 27, LDPR 21, New People 13. (confirmed, Wikipedia) - 2025-04-25: Russian CEC approves redistricting adding representation for annexed Ukrainian territories. (reported, Grokipedia) - 2025-05-23: Putin signs law creating 11 new single-mandate districts in the four annexed regions — first-ever single-mandate voting there. (reported) - 2026 (date unspecified pre-July): Putin decree formally sets election date as 20 September 2026 (voting spans 18–20 Sept). (confirmed, Interfax) - 2026-07-13/19: Anti-war candidate Boris Nadezhdin arrested, then ends campaign citing Kremlin pressure. (reported, Moscow Times) - 2026-08-10/11: Russia's only anti-war-leaning party barred from the ballot. (reported, Al Jazeera/Guardian/SCMP) - 2026-08-10: Moscow Times reports aggregated independent-leaning polling: ER 41%, CPRF 13%, New People 11%, LDPR 8%, SR 6% (well below ER's 2021 49.8% vote share). (reported) - 2026-08-17: Police detain opposition figures amid Supreme Court appeal over election ban. (reported, AA.com.tr) # Event Will United Russia gain the most seats (vs. its pre-election total) in the September 2026 Duma election? # Outcomes to forecast Yes / No # Kalshi market anchor No kalshi_direct data returned. The event ticker matches a Polymarket market currently priced at **69.5% YES** (7d: +2pp, 30d: +13pp; range 53.5–73.5% over 91 days; $6.88M volume) — treat this as the primary consensus anchor in absence of Kalshi order-book data. # Sub-question answers 1. **Polymarket price/history** — 69.5% YES currently, up sharply from a 53.5% low ~30-90 days ago, trending upward (+13pp/30d). (Polymarket direct) 2. **Current seat totals** — ER ~314-324 (sources vary due to byelections), CPRF 57, SR 27, LDPR 21-23, New People 13-15, other 4; sums to 450. ER's max possible gain is +126 seats (to 450); for CPRF to overtake ER on raw seats it would need +268. (Wikipedia, claude_news, code_execution) 3. **2021/2016 gain patterns** — 2011→2016: ER gained most (+105, rebound cycle). 2016→2021: CPRF gained most (+15), narrowly beating New People (+13, from zero base); ER *lost* 19 seats. Historical base rate: ER had the largest raw seat *gain* in only 1 of 3 recent cycles (33%), but held plurality of total seats in 3/3 (100%). (code_execution, Wikipedia) 4. **Election confirmed** — Yes, decreed for 18-20 Sept 2026 (Putin decree, Interfax). New 11 single-mandate seats added for annexed territories (formalized May 2025); total remains 450. No indications of postponement or martial-law delay found. 5. **Kremlin boosting ER vs. redistributing to satellites** — Mixed: Kremlin reportedly targets ER ≥70% support in occupied territories (Kyiv Post) suggesting ER seat gains there; but Nest Centre/Riddle report Kremlin is *shrinking* negotiated "quotas" for systemic opposition (10-13 seats vs ~20 in 2016/2021), which could paradoxically help or hurt ER's relative seat gain depending on how those seats reallocate. 6. **New/restructured party threat** — New People shows polling momentum (13.4% in some VCIOM readings, second place ahead of CPRF/LDPR in some polls) but from a small 13-15 seat base; no evidence of any party remotely approaching ER's total. No merger/veterans-party bloc large enough to outgain ER identified. # Key facts (high-confidence, factual) 1. [Wikipedia] ER won 324/450 seats in 2021 (down from 343 in 2016). 2. [Interfax] Election formally decreed for 20 Sept 2026 (voting 18-20 Sept). 3. [Grokipedia/claude_news] 11 new single-mandate seats added in annexed Ukrainian territories, first-ever vote there. 4. [Moscow Times] Independent-leaning polling shows ER support ~31-41%, down from 49.8% in 2021, but still a clear plurality over CPRF (~9-13%) and New People (~10-13%). 5. [claude_news] Kremlin reducing "cleared" constituency quotas for systemic opposition parties (10-13 vs ~20 previously). # Cross-market signals - Polymarket (same market): 69.5% YES, uptrending. - Related Polymarket markets: "Yabloko gain most seats" 0.25% YES; "A Just Russia gain most seats" 0.35% YES — both near-zero, consistent with ER as heavy favorite among named parties (implying ~99%+ probability collectively assigned to ER/CPRF/LDPR/New People bucket, with ER dominant). - No sportsbook data available. # Analyst opinions and speculation - Moscow Times/Nest Centre/BESA Center: consensus that ER will retain its constitutional supermajority; election is "for show," outcome not in doubt regarding ER's dominant position. - Kyiv Post/analysts flag Kremlin engineering of occupied-territory results (70% turnout/70% ER target), which mechanically pads ER's raw seat count via new districts. - Some analysts (Riddle) highlight New People/LDPR jockeying for "second place," but none suggest any party could approach ER's seat total. # Directional lean per outcome - **Yes (ER gains most)**: Supported by ER's total dominance, new annexed-territory seats likely to be ER-leaning, market consensus (69.5% and rising), and Kremlin's institutional control. Opposed by the resolution's literal "gain vs. before" framing — ER is near seat ceiling (324/450) with limited room to grow further, and in 2/3 recent cycles ER's seat count fell while an opposition party (CPRF) posted the largest gain from a depressed base. - **No (some other party gains most)**: Supported by historical pattern of mean-reversion after ER landslides (2021 cycle: CPRF gained most despite ER retaining plurality) and reduced quotas for opposition potentially reshuffling seats. Opposed by lack of any credible large low-base contender and near-total lack of political competitiveness in 2026 relative to 2021. # Gaps / unknowns - No Kalshi-direct order book data was returned for this ticker. - Ambiguity in whether "gain" is being priced by traders as literal seat-change vs. plurality-of-total (market behavior suggests traders may be treating it as the latter, understating "No" risk under strict resolution language). - No final seat allocation model or district-level ER performance data available closer to the vote. - Uncertain how newly created annexed-territory seats will be netted against 2021 baseline in official/consensus reporting. # Calibration anchors - Polymarket YES price (anchor): 69.5%, up from 53.5% low. - Base rate: ER held total plurality in 3/3 recent cycles (100%), but had the single largest raw seat *gain* in only 1/3 cycles (33%) — critical because this question's resolution criterion is explicitly gain-based, not total-based.

1. Both forecasters correctly flag the plurality-vs-gain ambiguity but resolve it very differently in magnitude (Forecast 1 blends 70/30 toward plurality reading yielding 0.81; Forecast 2 shades down to 0.62 without a clear explicit weighting model) — neither justifies why the *market itself* (traders on Polymarket) would resolve this using one interpretation over the other, despite the brief explicitly noting "market behavior suggests traders may be treating it as [plurality], understating 'No' risk under strict resolution language" — this is a critical uncertainty neither forecaster stress-tests with a sensitivity range. 2. Forecast 1's claim that "opposition parties losing rather than gaining seats" from shrunk quotas is asserted with high confidence but the brief itself says this is "Mixed... could paradoxically help or hurt ER's relative seat gain" — Forecast 1 overstates certainty on a genuinely ambiguous mechanism. 3. Neither forecaster deeply engages with the base-rate tension: ER's seat *gain* success rate is only 1/3 in recent cycles (33%), yet both land at 62-81% YES — implying they're leaning heavily on 2026-specific factors (repression, redistricting) to override the historical base rate, but neither explicitly quantifies how much weight the base rate deserves versus current-cycle idiosyncrasies, risking an anchoring/recency bias favoring the market price. 4. Neither considers a scenario where ER's official 2021 baseline (324) is compared against a post-redistricting 450-seat map where new annexed-territory seats are added on top — if reporting treats the *effective baseline* as 324+11=335 (new total seats) rather than net new seats being purely incremental gains, this could materially change what counts as "gain," a definitional edge case flagged in the "Gaps/unknowns" section but not modeled by either forecast. 5. Forecast 1's discounting "out of deference to the market's persistent discount" (0.85→0.81) is arbitrary and not tied to any specific evidence—it's essentially anchoring to the market price without independently justifying the magnitude of the haircut. 6. Neither forecaster addresses the possibility that CPRF, not New People, remains the most likely "gain" beneficiary based on the actual 2016→2021 precedent (CPRF gained +15 from a larger established base of 42, not from near-zero like New People's +13 from 3) — both discuss New People's momentum more than CPRF's structural advantage as the traditional second-place gain-taker, potentially underweighting the closest historical analog to a "No" scenario.

73% (52% confidence)

Yes: 73% No: 27%

The dominant uncertainty is definitional, not political: under the plurality reading ('most seats won'), ER is essentially certain (~98%) given ~41% polling versus ~11-13% for the nearest rival, a barred anti-war party, arrested challengers, and 11 new Kremlin-engineered annexed-territory districts. Under the strict raw-seat-*gain* reading, however, ER is actually disfavored — its list share falling from 49.8% to ~41% implies flat-to-declining totals from a near-ceiling 314-324 base, while New People (5.3%→11-13%) or a resurgent satellite could post the largest raw gain, mirroring the 2016→2021 pattern where CPRF gained most while ER lost 19 seats; that branch is maybe 35-40% YES. Blending roughly 65/35 toward the plurality reading yields ~0.76, and I shade down slightly toward the $6.9M-volume Polymarket anchor of 69.5% — a price that, given near-zero quotes on Yabloko/SR and ~30% distributed elsewhere, suggests informed traders assign non-trivial weight to the strict-gain mechanics. Forecast 1 (0.81) overweights the plurality reading and overstates certainty on the ambiguous opposition-quota mechanism; Forecast 2 (0.62) under-weights how loosely 'gain the most seats' is typically resolved. Final: 0.73.

Total pipeline time: 351.9s