Analysis
Neither bloc reaches 61. Two together do.
Read as two blocs, this map has no majority in it. Almost nobody reads it as two — and counted the way most Israeli analysis counts it, the anti-Netanyahu side clears 61 in nine simulations out of ten.
Written on 21 August 2026, and the figures below are that day’s. Two days later the forecast has the Netanyahu bloc with 54 seats, the Opposition bloc with 55 and the Arab parties with 11; the two blocs together reach 61 in 92% of simulations, the opposition alone in 16% and the Netanyahu bloc in 5%. The four-seat gap this piece is about has closed to one, which is the argument rather than an objection to it.
The seat forecast for 27 October has the Opposition bloc on about 56 seats and the Netanyahu bloc on about 52. Neither reaches 61. More to the point, neither is likely to: the Opposition bloc clears a majority in 19% of simulations and the Netanyahu bloc in 5%.
That is not a close call being reported cautiously. Read as two blocs, it is a map in which the most probable outcome is that neither side can govern alone.
Almost nobody reads it as two.
The eleven seats a two-bloc reading leaves out
The Arab parties are forecast at 11 seats, held as a third bloc rather than folded into the opposition. Add them and the arithmetic changes completely:
| Median | 90% range | Reaches 61 | |
|---|---|---|---|
| Netanyahu bloc | 52 | 43–60 | 5% |
| Opposition bloc | 56 | 47–65 | 19% |
| Opposition + Arab parties | 67 | 59–76 | 91% |
Ninety-one per cent is not a close call either, and it points the other way. This is why the prevailing read of the 2026 map is that the anti-Netanyahu side is ahead rather than stuck, and any analysis that stops at 56 against 52 has told you the least interesting half of it.
We keep the three blocs apart anyway, for one statistical reason and one political one, and they are worth separating. The statistical reason: the error model lets seats move freely inside a bloc, and seats do not move freely between Yesh Atid and Hadash-Ta’al. Folding them together would model a transfer that does not happen, and would report bloc probabilities that are too confident.
The political reason is that those 11 seats are not one actor. Ra’am entered a governing coalition in 2021 under Mansour Abbas — the first independent Arab party to do so. Balad has ruled it out, and this forecast gives it a 60% chance of winning no seats at all. “Opposition plus Arab parties” is a sum, not a bloc that has agreed to anything.
So 67 is arithmetic. It is the number a reader should have, and it is still not a prediction that a government forms — for the reason the last section of this piece is about.
Why the four-seat gap is not the story
A 4-seat lead sounds like a lead. Between the two major blocs it is worth almost nothing, because the question Israeli coalition arithmetic asks is not “who is ahead” but “who reaches 61”. A bloc on 56 and a bloc on 52 are both losing to the same threshold, and the four seats between them only matter alongside the eleven above.
The place to look is the parties on the line. Five lists sit at real risk of winning no seats at all — Balad at 60%, Unity at 60%, the residual “Others” bucket at 60%, Blue & White at 59% and Zionist Home at 39% — and a list that falls below the 3.25% threshold wins nothing: no other list receives its votes, and they are left out of the allocation. Those are the outcomes that move the seat map, far more than a point of national polling.
The precedent nobody should forget
April 2019 is the standing reminder that 61 seats and a government are different things, and it is worth being precise about why, because at first glance it looks like a contradiction.
A bloc is a count of the lists that endorsed a candidate for prime minister. It is not a coalition agreement. After the April 2019 election, 65 MKs formally recommended Netanyahu to President Rivlin — Likud 35, United Torah Judaism 8, Shas 8, Yisrael Beiteinu 5, the Union of Right-Wing Parties 5 and Kulanu 4 — and on that basis he was tasked with forming a government.
He could not. Yisrael Beiteinu had recommended him and still would not enter the coalition: Avigdor Lieberman insisted the Haredi conscription bill pass in its original form, and Shas and United Torah Judaism, 16 seats between them, would not have it. Take Yisrael Beiteinu’s five away and Netanyahu was on 60. One short.
What followed had not happened before. Rather than let the mandate pass to Benny Gantz, the Knesset voted on 30 May to dissolve itself, and the country went back to the polls in September.
So 65 MKs wanted him as prime minister and no 61 of them could agree on terms. Both things are true, and only the first is a number a seat forecast can produce.
So read the number for what it is. This model forecasts how seats are likely to fall. It does not model coalition bargaining, party vetoes, or who is willing to sit with whom, and a bloc reaching 61 here still has to go and agree a coalition.