Israel’s pollsters cannot agree who is leading, and agree nobody can form a government
Channel 13 gives Netanyahu’s bloc 55 seats, Channel 12 gives Eisenkot’s 54, Kan called it 52–52. The spread between houses is wider than any margin of error, and the decisive number is one nobody polls.
Channel 13’s poll, published Wednesday evening and carried in ToI’s Thursday liveblog, gives the Netanyahu bloc 55 seats and the Zionist anti-Netanyahu bloc 50, with Likud and Gadi Eisenkot’s Yashar “neck and neck.” It has Otzma Yehudit and United Torah Judaism on 8 each, Shas 7, Religious Zionism–Zehut 7, Am Yisrael 5; Naftali Bennett’s B’Yachad 12, The Democrats 10, Yisrael Beytenu 8; Ra’am 6 and the Joint List 9. The Reservists and Blue and White fall below the 3.25% threshold. Sample 1,013, margin of error 2.5%. ToI’s own framing is that Netanyahu is still not predicted to be able to form a coalition. Channel 12, two days earlier, had the mirror image: Eisenkot’s bloc 54 to Netanyahu’s 49 — and short of a majority without the Arab parties.
The houses diverge by more than sampling noise. Kan’s late-September poll had the blocs tied 52–52, Yashar on 23 and Likud 21, with Ofer Winter’s four seats carrying Netanyahu to 52, and a preferred-prime-minister split of 39% to 38%. A JPost roundup documents i24NEWS/Direct Polls putting Likud first on 27 in late September while Kan, Maariv, N12 and Channel 13 all had Yashar ahead. The Channel 12 series runs through Mano Geva’s Midgam Institute with iPanel — 501 respondents drawn from 2,498 approached, margin of error 4.4%. A Channel 13 poll on 1 October had Netanyahu 54, the opposition 52 and the Arab parties 14, which Haaretz read as the finding that matters: the Zionist opposition’s refusal to work with Arab lists is what keeps handing Netanyahu the edge.
Assessment: Treat bloc totals as an artefact of who the pollster counts, not a forecast. The 14 Arab-party seats are the swing variable, and they are excluded by convention rather than arithmetic — which means the election can be lost by the opposition after the votes are counted. The second mechanism to watch is the 3.25% threshold: Winter, Gantz’s Blue and White, Noam, the Haredi Public list and the Reservists are all hovering, and every list that fails redistributes its seats to the largest parties in its own camp. A two-seat threshold accident is worth more than a week of campaigning. Results are due 4 November; coalition arithmetic will run well past it.