Scan: Will Likud win fewer than 20 seats in the 2026 Israeli legislative election?

1. Planner Fable 5.1 · Fermi decomposition + tool plan

Tools chosen: polymarket_direct polymarket_related kalshi_related article_search claude_news gdelt_news wikipedia code_execution

Sub-questions (Fermi decomposition)

What is the current Polymarket price for Likud <20 seats and how have the sibling seat-bracket markets (20-24, 25-29, 30+) been priced?
What are the most recent Israeli opinion polls (Channel 12, Channel 13, Kan, Maariv) showing for Likud's projected seat count, and what is the trend over the past 6 months?
Is there any credible prospect of Likud splitting (e.g., a Netanyahu leadership challenge, defections to Bennett/Eisenkot/Liberman) or merging with another party before the election, which would change the counted seat total?
Has the election date been officially set, and is there any chance results are not definitively known by Oct 31, 2027 (triggering resolution to the lowest bracket)?
Historically, how large has the gap been between final-year polling and actual Likud results in Israeli elections (polling error magnitude), and how often has Likud fallen below 20 seats?
What is Netanyahu's political status (corruption trial, health, pardon talks, coalition stability with Haredi parties over the draft law) and could it drive a mass voter shift away from Likud?

Planner reasoning

Likud currently holds 32 seats and polls have generally shown it in the mid-20s to low-30s, so 'fewer than 20' is a tail outcome that hinges on whether Netanyahu's coalition collapses further, whether Likud splits or merges, and how polling error behaves. Key drivers: current polling aggregates, any leadership challenge or splinter, election timing, and historical Likud floor (lowest ever was 12 seats in 2006 after Kadima split). I'll anchor on the Polymarket price, check sibling seat brackets and Kalshi for cross-market signals, and pull recent polling news.

2. Research Parallel tool calls

tool	status	time	items	summary
polymarket_direct	OK	0.5s	1	## This Market's Polymarket Data Will Likud win fewer than 20 seats in the 2026 Israeli legislative election? - Current price (probability): 14.80% - 7-day price change: -0.90% - 30-day price change: -4.85% - Total volume: $50,061 (USD notional) - Price range: 9.40% - 25.80% - Data points: 90 d
polymarket_related	OK	2.8s	0	Scanned 100 active Polymarket markets, kept 0 matches. keyword 'Likud': 0 markets \| keyword 'Israeli election': 0 markets \| keyword 'Netanyahu': 0 markets \| keyword 'Knesset seats': 0 markets
kalshi_related	OK	2.7s	2	2 related markets / summaries. keyword 'Israel election': ok \| keyword 'Likud': no matches \| keyword 'Netanyahu': ok
article_search	OK	47.1s	20	Fetched 20 unique articles across 4 queries (lookback=45d). 'Likud poll seats Knesset': 10 articles \| 'Israel election polls Netanyahu Likud': 10 articles \| 'Likud split defection Netanyahu challenge': 10 articles \| 'Israel 2026 election date Knesset dissolution': 10 articles
claude_news	OK	24.9s	8	## Findings - Election date confirmed: The coalition has officially set the 2026 Israeli election for October 27, 2026, marking the first time since 1988 that a Knesset will complete a full term ([Israel Hayom](https://www.israelhayom.com/2026/07/12/israels-next-official-election-date-set/)). M
gdelt_news	OK	185.5s	10	GDELT: 10 articles across 3 queries (lookback=30d). 'Likud poll seats': error GDELT rate-limited after retries (429) \| 'Netanyahu Likud election': 10 hits \| 'Israel election Knesset polls': error GDELT rate-limited after retries (429)
wikipedia	OK	0.1s	3	Fetched 3 Wikipedia entries (0 missing pages).
code_execution	OK	33.1s	0	## Findings: Likud <20-Seat Probability Model (2026 Election) Historical polling error (1999–2022, 11 elections): - Mean error (actual − final poll avg): +0.36 seats (slight positive bias — Likud has modestly outperformed polls on average) - Raw std dev of error: ~5.1 seats, driven heav

3. Evidence Brief Sonnet 5 · 7029 chars

# Current state Israel's 2026 Knesset election is confirmed for October 27, 2026 — the first full-term election since 1988. Likud's seat projections are unusually volatile and fragmented across pollsters, ranging from 19 to 27 seats depending on the poll, straddling the market's 20-seat threshold. No merger or split has occurred; party lists have closed as of early September 2026. # Timeline of key events - 2022-11-01: Likud won 32 seats in prior election (confirmed baseline, Wikipedia). - 2026-07-12: Coalition officially sets Oct 27, 2026 as election date (confirmed — Israel Hayom, IFES, Britannica). - 2026-08-17/18: Likud holds internal primaries; polls at the time show Likud ~22 seats, several sitting MKs at risk (reported — Haaretz). - 2026-08-24/25: Netanyahu escalates rhetoric vs. Turkey amid campaign framed around security; critics allege political motive given trailing polls (reported — CNN). - 2026-08-26/27: Channel 13 poll shows Likud sinking to 19 seats (a 4-seat drop); parallel reports of Netanyahu in closed talks warning "Likud will lose unless situation changes" and weighing a rotation government with Eisenkot (reported/rumored — Haaretz, Middle East Monitor, JPost). - 2026-08-31: Ben Gvir publicly rejects Netanyahu's push to merge Otzma Yehudit with Religious Zionism; far-right list fragmentation persists, with new parties (Winter's "Your People Israel") entering the field (confirmed — Times of Israel, CNN). - Early Sept 2026: Knesset candidate lists close; no Likud merger occurs. Opposition forms surplus-vote agreements (Bennett-Liberman, Eisenkot-Golan); Likud reportedly pairs with Religious Zionist-Zehut for surplus votes (confirmed — JPost, Israel Policy Forum). - 2026-09-09: i24NEWS/Direct Polls poll shows Likud rebounding to lead with 27 seats — an outlier vs. other concurrent polls (reported — i24NEWS). - ~2026-09-10: Post-list-closing wave of 5 polls shows fragmented picture; Yashar (Eisenkot) leads in 4 of 5, Likud leads only in the i24 outlier poll (reported — JPost). # Event Will Likud win fewer than 20 seats in the 2026 Israeli legislative election? # Outcomes to forecast Yes (Likud <20 seats) / No (Likud ≥20 seats) # Kalshi market anchor No distinct Kalshi-direct price was returned for this ticker; the only direct pricing data available for this exact market came via the polymarket_direct tool (same ticker ID format), showing **14.8% current price** for "Yes" (<20 seats), down 0.9% over 7 days and down 4.85% over 30 days. 90-day range: 9.4%–25.8%. Volume modest ($50K total). Treat 14.8% as the best available consensus anchor pending confirmation of a separate Kalshi orderbook price. # Sub-question answers 1. **Polymarket pricing for <20 and siblings** — Only the <20 bracket was directly retrieved (14.8%, trending down over 30d). No sibling 20-24/25-29/30+ Polymarket data surfaced (polymarket_related found 0 matches), so cross-bracket consistency can't be verified. 2. **Recent polls/trend** — Wide dispersion: Channel 13 (late Aug) 19 seats; primary-period polls ~22; most other polls 25-27; i24/Direct Polls (Sept 9) 27 (outlier). Overall down from 32 in 2022 due to judicial overhaul fallout, Oct 7 war, coalition crises (Israel Policy Forum). 3. **Split/merger risk** — No Likud split reported; no internal leadership challenge to Netanyahu found. Netanyahu tried to broker a Ben Gvir–Smotrich merger; Ben Gvir refused. Lists have closed with Likud running solo (surplus agreement with Religious Zionist-Zehut only) — merger/split scenario effectively foreclosed for this cycle (Times of Israel, JPost). 4. **Election date certainty** — Confirmed for Oct 27, 2026 via multiple independent sources; no indications of postponement risk. The 2027 fallback-to-lowest-bracket clause appears unlikely to trigger. 5. **Historical polling error** — Code-execution model of 1999-2022: mean error +0.36 seats (slight actual-outperformance vs. polls), std ~5.1, driven by 1999 collapse (poll ~30→actual 19). Excluding 1999, typical error ±2-6 seats. Likud has fallen below 20 seats only once historically (1999). 6. **Netanyahu status** — Reports describe a struggling campaign, internal alarm ("Likud will lose unless situation changes"), and Netanyahu weighing a rotation government with Eisenkot as a fallback — signals of a genuinely weak position, though no confirmed mass defection or corruption-trial resolution/pardon reported in this research window (JPost, Middle East Monitor, Haaretz). # Key facts (high-confidence, factual) 1. [Wikipedia] Election legally set for Oct 27, 2026; 3.25% threshold applies. 2. [Israel Policy Forum] Likud down from 32 (2022) amid judicial overhaul/Oct 7 fallout. 3. [Haaretz] Channel 13 late-Aug poll: Likud 19 seats. 4. [i24NEWS] Sept 9 outlier poll: Likud 27 seats, largest party. 5. [JPost] Post-list-closing polls (5 surveys): Yashar leads in 4/5; Likud leads only in i24 poll. 6. [Times of Israel] Ben Gvir rejected Netanyahu's Smotrich-merger demand; no Likud merger occurred. 7. [Polymarket] <20-seat Yes price = 14.8%, declining trend. # Cross-market signals - Kalshi related: Israel Defense Minister market (Katz, 11%, falling sharply) and Next PM market (Bennett, 10%) both suggest markets pricing meaningful odds of Likud/Netanyahu underperformance, though not seat-specific. - Polymarket: 14.8% Yes for <20 seats; no sibling bracket data retrieved. - Sportsbook implied: none available. # Analyst opinions and speculation - Some Netanyahu critics (Barak, Golan) allege security escalations are politically motivated ahead of trailing polls (CNN). - Reports of Netanyahu weighing a rotation government with Eisenkot suggest internal expectation of a weak Likud result, though this is speculative/rumored. # Directional lean per outcome - **Yes (<20 seats)**: Supported by one low outlier poll (19), a documented downward trend from 2022, historical precedent of large swings (1999 collapse), and internal Likud alarm reports. Opposing: most concurrent polls (22-27) sit above threshold; model estimates ~9-17% probability if poll average is mid-20s. - **No (≥20 seats)**: Supported by majority of polls clustering 22-27, no merger/split materializing, confirmed election timeline, and Polymarket pricing only 14.8% for Yes (implying ~85% for No). # Gaps / unknowns - No confirmed Kalshi-direct order book price distinct from the Polymarket-style figure supplied. - No sibling-bracket Polymarket data to cross-check consistency/arbitrage. - Netanyahu's corruption trial status/pardon talks not detailed in retrieved research. - Poll aggregator average (true current consensus) not explicitly computed — range 19-27 with no clear median stated. # Calibration anchors - Current market (Polymarket-style) YES price: **14.8%** (declining). - Historical base rate: Likud has fallen below 20 seats once in ~11 elections (1999), ~9% historical frequency, consistent with current pricing. - Quant model: ~9-17% probability if poll average mid-20s (24-26), matching market's ~15% anchor reasonably well.

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

claude-fable-5-1

52% conf

Yes	15%
No	85%

The market anchor is ~14.8% for Yes with a declining 30-day trend, and the brief's evidence largely corroborates it: most concurrent polls place Likud at 22-27 seats, with only one Channel 13 outlier at 19 and an i24 outlier at 27, implying a poll average in the low-to-mid 20s. Historical Israeli polling error (std ~5 seats, slight Likud outperformance on average) applied to a ~24-seat average gives roughly a 10-18% chance of finishing below 20. Downside risks exist — internal Likud alarm, Netanyahu weighing a rotation with Eisenkot, and the 1999-style collapse precedent — but lists have closed with no split, and Likud typically consolidates right-wing votes late in campaigns. I stay close to the market at 0.15 for Yes.

gpt-5.6-sol

68% conf

Yes	15%
No	85%

The 14.8% market price for Yes is the best available consensus anchor and aligns with both the roughly 9% historical base rate and the brief's 9-17% model range. Most polls place Likud at 22-27 seats, pointing toward No, although Channel 13's 19-seat result shows that Yes remains plausible amid substantial pollster dispersion. Closed candidate lists and the absence of a Likud split remove a major route to falling below 20, while internal alarm and campaign weakness provide only a modest counterweight. I therefore stay close to the market at 15% Yes.

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

Both forecasters converge nearly identically on 15% with minimal independent reasoning, essentially anchoring on the Polymarket price without stress-testing it — e.g., neither notes the price has been declining 30 days (25.8%→14.8% over 90 days) even as post-list-closing polls showed 4/5 surveys with Yashar leading and Likud only winning the outlier i24 poll, which could argue for a *higher* Yes probability than the stale 14.8% anchor reflects.

Neither forecaster grapples with the "poll average" ambiguity flagged in the brief — the Gaps section explicitly states "no clear median stated," yet both assert an implicit mid-20s average by weighting the 22-27 cluster over the 19-seat Channel 13 poll and treating i24's 27 as a legitimate data point despite it being explicitly labeled an "outlier" possibly to be discounted more heavily than Channel 13's 19.

Both treat the 1999 collapse (poll ~30→actual 19) as merely providing historical std/tail risk, but neither considers that current conditions (Oct 7 war fallout, judicial overhaul backlash, corruption trial, internal "rotation government" talk) more closely resemble a genuine confidence-crisis scenario than a normal-variance year — arguably justifying fatter left-tail weight than the symmetric ±5-seat model implies.

Neither forecaster addresses the cross-market signal that Kalshi's Defense Minister market (Katz, 11%, "falling sharply") and Next PM market (Bennett, 10%) both imply markets are pricing meaningful Likud/Netanyahu underperformance — this could be corroborating evidence for a higher Yes probability that neither analyst incorporates into their reasoning beyond a passing mention.

Both forecasters lean heavily on "closed lists foreclose merger/split" as evidence for No, but this cuts both ways: with Ben Gvir refusing to merge and far-right fragmentation persisting (Winter's new party entering), right-wing vote-splitting could suppress Likud's own surplus-vote efficiency even without an internal Likud split, a channel neither forecaster explores as a distinct downside risk beyond noting the Zehut surplus pact.

Given confidence scores of 0.52 and 0.68 attached to identical 15%/85% splits, there's an inconsistency in calibration language — Forecast 2 claims higher confidence despite acknowledging the same wide poll dispersion (19-27) and same anchor, suggesting the confidence metric isn't tracking actual evidentiary uncertainty.

6. Aggregator Fable 5.1 · final calibrated forecast

17% (60% confidence)

Yes: 17% No: 83%

Both forecasters anchor on the ~14.8% market price, which is consistent with a poll cluster of 22-27 seats, a ~9% historical base rate for Likud falling below 20, and Likud's tendency to consolidate right-wing votes late and slightly outperform polls. The critique is partially valid: the 19-seat Channel 13 poll, internal Likud alarm ('will lose unless situation changes'), rotation-government talk, and the confidence-crisis conditions resembling 1999 argue for a somewhat fatter left tail than a symmetric error model, and the i24 27-seat poll should be discounted as an outlier. However, Yashar leading in 4/5 polls does not imply Likud below 20 (it can lead with Likud at 22-25), lists have closed without a split, and the market has been drifting toward No over 30 days rather than away. I nudge slightly above the anchor to 17% Yes to reflect the tail risk with seven weeks of volatile campaign remaining.

Pipeline Timing

Total pipeline time: 294.2s

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