Israel Election 2026 Guide: Election Math and How to Read the Polls

Why seats matter more than votes

Israeli polls report seats out of 120, not vote share. That is the number that decides who governs, because a government needs 61. Vote share only matters at one point: whether a party clears 3.25%.

Polls also disagree with each other, and the gaps are large. In the nine seat polls published between 24 September and 7 October, Likud ranged from 18 seats to 31. The right way to read them is as a range, not a headline. This guide shows how.

One note on numbers across this site. The party profiles quote a simple average of the most recent polls, from 24 September to 7 October. The poll tracker averages each pollster first and then the pollsters, using every qualifying poll since 15 September. The two methods give slightly different figures. Neither is a forecast.

The camps

Israeli media sort parties into camps. Nobody wins 61 alone, so the question is which camp can get there.

Some outlets add Winter's Amcha Yisrael to the coalition, which lifts it to about 53.5. Winter says he will recommend only Netanyahu. Hendel says his party belongs to neither camp.

Neither main camp reaches 61 in most polls. The coalition runs about 49 to 55. The Zionist opposition runs about 50 to 55. Only Channel 14 shows the coalition above 61, at 63.

The opposition has three ways to 61. It can add the Arab parties, add Hendel's four seats, or hope that a party near the threshold fails. Add the Arab parties and the anti-Netanyahu total runs about 63 to 67 in most polls. But Bennett and Hendel say they will not rely on Arab parties, while the Democrats want Ra'am in. An earlier Channel 12 poll found most opposition voters would accept a government that an Arab party supports from outside.

Labels are not loyalty. Shas and UTJ sit in the coalition bloc on paper. Maariv's Haredi sources say they plan to sit in opposition for the first year, and Eisenkot says he would sit with Shas. Treat the bloc totals as a starting point.

How to read a seat poll

A seat poll is a small survey turned into seats by a model. Four things to check.

• Sample size. Most Israeli polls this cycle sample 500 to 600 people. With a sample that size, one respondent can be the difference between a party at 4 seats and 0.
• Margin of error. About 4 points on the full sample, and much wider for any single party. A move of one or two seats between polls is usually noise.
• Fieldwork date. Polls are published a day or more after fieldwork. Use the date the survey was run, not the headline date.
• The pollster. Each outfit has a house effect, a habit of running high or low for certain parties. The biggest gap this cycle is Channel 14, which shows Likud at 27 to 32 seats while Channels 12 and 13, Kan and Maariv show it at 18 to 21. Other pollsters and analysts treat Channel 14 as the outlier. Be fair about its record, though. In 2022 its pollster, Direct Polls, was closest on the size of the right-wing bloc, 62 against a real 64, even as it overstated Likud.

The practical advice is to average across pollsters and track the trend, not any one poll. Throwing out the highest and lowest outlier is a sensible start.

The threshold in the polls

Pollsters publish percentages for small parties, and the 3.25% line is where the story is. A party polling 3.0% to 3.5% is close to a coin flip. A party polling 1% is not in the race.

The same party can look very different from one poll to the next:

• Winter's Amcha Yisrael was at 1.5% on Channel 12 and 5 seats on Channel 13 in the same week.
• Hendel and Zelekha were at 4 seats in most polls, but 2.9% on Channel 13 on 7 October and 1.8% on Channel 14.

Here is how one party's fate changes the math.

1. Hendel and Zelekha pass. Their 4 seats are unaligned. Maariv said they could take the opposition to 61 without the Arab parties. They also say they will not complete a majority with the Haredim.
2. Hendel and Zelekha fail. Their votes are wasted. The other lists gain about 3% more seats in proportion, so each side picks up a seat or two.
3. Winter fails. His 3 to 4 seats are redistributed. The coalition loses more than the opposition because Winter's voters lean right, and Winter's own seats would have counted for Netanyahu.
4. An Arab list falls under the line. The Joint List polls near 7 seats, about 6% of the total. If it failed, that 6% would be wasted and every surviving list would gain about 6% more seats. A bloc on 50 would get about 53. That can flip a close race. The Supreme Court's 2 October rulings removed the immediate disqualification risk, but turnout still matters.

The lesson: when you read a bloc total, ask which near-threshold parties are inside it.

Where the polls stand

Simple average of nine seat polls published between 24 September and 7 October 2026 (Kantar/Kan, Maariv/Lazar, Direct Polls/i24, Channel 12, Channel 13, Channel 14). A poll that shows a party under the threshold counts as 0 seats. Averages do not add to 120 for that reason.

Parties well under the line: Blue and White (Gantz) at 0.5% to 1.3%, Haredi Public Party (Leitner) at 0.4% to 1.1%, and Noam (Maoz) at 0.2% to 0.9%.

Two cautions. First, Likud's average is pulled up by Channel 14 and i24. Without them it is about 19.7. Second, the published bloc totals vary: coalition 49 to 63, Zionist opposition 45 to 55, Arab parties 12 to 15. The Middle East Forum's 14-day average puts the Netanyahu bloc at 53, the opposition at 54 and the Arab parties at 13.

What the track record says

Israeli polls get the big picture right and the edges wrong.

• 2022 Meretz. All ten polls in the final week gave Meretz 4 to 5 seats. It finished at 3.16% and won none. Balad (2.91%) and Jewish Home (1.19%) were correctly polled under the line. About 289,000 anti-Netanyahu votes were wasted.
• 2022 blocs. Final polls put the Netanyahu bloc at 59 to 62 seats. It won 64. Likud was polled at 30 to 34 and won 32.
• April 2019. Bennett's New Right polled above the threshold and missed it, with 3.22% against the 3.25% line, about 1,400 votes short. It won no seats.

Take from this that a party polling at 4 seats is a bet, not a result. Bloc totals tend to be closer than single-party numbers, but errors of two to four seats on a bloc happen, and 61 is a hard line.

A checklist for any Israeli poll

1. Who ran it, and for which outlet? Compare with that pollster's own past polls.
2. What were the fieldwork dates?
3. How big was the sample? Under 600 means small parties are noisy.
4. Is each small party shown as seats or as a percentage? Check the percentage against 3.25%.
5. Which parties are near the line, and which bloc do they sit in?
6. Are Arab parties counted in a bloc, or listed apart?
7. Is Shas and UTJ's place in the coalition bloc a label or a commitment?
8. How does this poll compare with the average of the last week?

Polls cannot be published from the end of the Friday before the election, 23 October, until voting ends on 27 October.

Sources

• Middle East Forum poll of polls
• JPost: Maariv poll, 30 Sep to 1 Oct
• JPost: Maariv poll, 23 to 24 Sep
• JPost: poll roundup, 27 Sep
• Israel National News: Channel 13 and 14, 7 Oct
• Israel National News: Kan, 4 Oct
• Israel National News: Channel 12, 6 Oct
• JPost: bloc analysis, 8 Oct
• Wikipedia: 2022 Israeli legislative election
• Wikipedia: opinion polling for 2022
• Times of Israel: how to make sense of the polls
• JPost opinion: polling gaps