Amcha Israel: no seats, six points on the bloc

Ofer Winter’s list, Amcha Israel, does not appear in the chamber on this page. Its median is 3 seats. The electoral threshold, at 3.25% of the vote, does not allow a list fewer than 4, and the allocation this page prints gives a list nothing when its median is a number the threshold forbids. So it is drawn as zero.

That does not make it a small row. Holding it inside the Netanyahu bloc rather than outside it takes that bloc’s chance of reaching 61 seats from 2.0% to 7.9% — the same draws, the same list, and nearly six points of difference in the Netanyahu bloc’s headline probability, decided by one line in a mapping file.

This is about how that line got written, and why a list a reader cannot see is worth reading carefully.

The field is split down the middle

Twelve polls in the model’s window carry a reading for this list. They divide exactly in half, and not by pollster.

Six report it as a percentage, which is what a poll does when a list falls short: 0.6%, 0.9%, 1.5%, 2.1%, 2.2% and 2.5%, against a threshold of 3.25%.

Six report it with seats: 4, 5, 6, 6, 7 and 8.

The division is by date. Every percentage was published on 26 August or earlier; every seat figure on 26 August or later.

The pivot day is worth stopping on, because on it the field disagreed with itself. Four firms published on 26 August. Direct Polls for i24 News put the list on 8 seats, Maagar Mochot for Channel 16 on 7, Lazar for Walla on 6 — and HaMadad for Channel 13, working the largest sample of the four at 1,407, had it at 2.2% and out of the Knesset entirely. One day, four pollsters, and no agreement about whether the list would be in the Knesset at all.

A percentage is weaker evidence than it looks

The obvious reading of six sub-threshold percentages is that they bracket the list neatly below the line. They do not, and the reason is that they are only ever published while the list is losing.

A pollster prints a share for a list that fell short and prints seats for a list that cleared. The corpus therefore holds six measurements of how far short this list fell and not one measurement of how far over it went. The evidence is censored at exactly the point the question turns on.

That is why the run page marks this list partial rather than measured, and why the model gives it a 90% range of 1 to 6 seats around a median of 3 — a span running from a list that misses the threshold to one with a caucus of six. It gives the list 4 seats or more, the fewest the threshold allows, in 48% of simulations, and gives it nothing at all in 5%. That is not the model hedging. Those are both things the evidence supports, at the same time.

The bloc was decided twice

For almost every list, bloc membership is not a decision anyone here makes. It follows the coding published with the poll compilation, which is the standard Israeli media framing.

Amcha Israel was the exception. The compilation coded it unaligned — the only 2026 list it declined to place — while TheMadad coded it coalition. The two sources that code blocs disagreed, and this project’s own rule for breaking such a tie, which asks whose electorate a list’s votes move freely with, had nothing to work on: every reading was below the line.

On 27 August this project put the list in the Netanyahu bloc anyway, on the owner’s judgement, and wrote that into the mapping file marked as an exception rather than folded in quietly — with a note saying to revisit it as soon as there was something to measure it against.

There was, that same evening. In a video reported that night by Ynet, Maariv, JDN and Kipa, Winter said:

אנחנו במחנה הימין ומנהיג מחנה הימין הוא נתניהו. רק עליו נמליץ

We are in the right-wing camp, and the leader of the right-wing camp is Netanyahu. We will recommend only him.

That is not a hint about a coalition. A public endorsement of Netanyahu for prime minister is the criterion the bloc file is built on, stated in its own terms by the person entitled to state it.

Two days later, on 29 August, the poll compilation recoded all eleven of its Amcha rows from unaligned to pro_netanyahu.

One thing is worth being exact about, because it would be easy to report as a change and it is not one: nothing in this pipeline reads that column. Bloc membership comes from the mapping file and nowhere else. The recoding moved no number here. It is a second source arriving at the answer we had already written down, which is all it is and all we asked of it.

Whose votes these are, and what we do not claim about them

N12’s profile of Winter, published on 28 August, is the fullest account so far of the constituency this list is built for. Its argument is that the old alliance between the knitted-kippah national religious and the Haredi parties has come apart over the Haredi refusal to serve, and that the party holding the name Religious Zionism sided with the Haredim. Avi Benayahu, a former IDF spokesman, puts it there as הרבה מאוד אנשים בציונות הדתית מרגישים הומלסים — a great many people in religious Zionism feel homeless.

The same piece records that the alignment is contested by people watching it closely. Netanyahu had offered Winter a safe place on the Likud list and Winter did not take the call; Netanyahu then publicly urged voters not to vote for him. The commentator Ravit Hecht reads the whole quarrel as staged and calls Winter Netanyahu’s perfect tax shelter — a way for right-wing voters who cannot bring themselves to back Netanyahu to support the coalition without saying so. Benayahu, in the same piece, will say only that Winter is on the right and that nobody can be sure what he does after the ballot.

None of that is in the model, and this page will not pretend otherwise. What happens after an election is coalition bargaining, which this project does not forecast. The bloc file records a declared position before the ballot, and that is the whole of what it records.

One thing is in the model. Religious Zionism sits at a median of 4 seats — exactly what the threshold allows, with nothing to spare, winning nothing in 3.6% of simulations. It and Amcha Israel are read together on every round, not because a theory says one is taking the other’s voters, but because two lists this close to the line and this close to each other cannot be assessed apart.

What moved this round

Two polls arrived: Filber for Channel 14 on 27 August, on a sample of 981, and Lazar for Maariv, fielded 26–27 August on a sample of 506. Both give Amcha Israel seats — 4 and 5 respectively.

No median on this page changed. Likud and Yashar remain level with 23 seats apiece; Likud finishes largest in 52% of simulations against Yashar’s 48%, which is to say the order at the top is not settled and should not be read as though it were.

The three blocs read 52 for the Netanyahu bloc, 57 for the opposition and 11 for the Arab parties. As always those are sums of seats under a published bloc coding, not a forecast of who will govern. The opposition and the Arab parties hold 68 between them, which is arithmetic and nothing more: it is six separate lists that do not act as one, and nothing here suggests they would sit together.