Who stays home on election day?

Participation by list

Would full turnout change the government?

The one result that holds across methods: the Arab lists gain. Arab-majority ballot boxes vote at 43–64%, against 62–68% in general boxes. Under Hypothetical 1 the Arab lists gain 1–6 seats in every election. From 2019 on, those seats come mostly out of the Netanyahu bloc: in 2022 it falls from a majority to 60 seats.

Exercise: only non-Haredi Jewish supporters turn out

The scenario above raises turnout for everyone. Here only the supporters of non-Haredi Jewish lists are brought to 100% (Likud, the centre, the left, the religious right, Yisrael Beiteinu). The Arab lists and the Haredi lists (Shas, UTJ) keep exactly the votes they got. Think of a mobilisation campaign aimed only at the Jewish non-Haredi public.

Where the turnout gap comes from

Haredi boxes always vote most (69–75%). Arab-box turnout swings the most: 43% in 2021, when the Joint List split, and 64% in 2020, when the united list won 15 seats. In general boxes turnout climbs with the neighbourhood's socio-economic cluster, from about 50% to 71%. Arab and Haredi boxes show no such climb.

Fifteen years of participation rates

Lists come and go, so they are grouped into families. Each cell is the family's votes ÷ its potential supporters.

Voters living abroad

Israel has no postal voting, and citizens who move abroad stay on the register. So every ballot box carries some "phantom" voters who could not vote even if they wanted to. The CBS publishes a national estimate before each election:

Taking them out raises 2022 turnout among people actually living in Israel from the official 70.6% to about 77%. Abroad voters are not spread evenly. The census shows where they are: compare each town's registered voters with its adult residents in the 2022 census.

Expats flying in to vote (Knesset 26, 27 Oct 2026)

Israelis abroad can vote only by flying home. Ahead of the 2026 election this became an organised effort. The AID Coalition's Fly&Vote project estimates 53,000–67,000 expats will come. Its survey of about 23,000 Israelis abroad found 80% intending to travel (48% from the US, 37% from Europe). Only 14% of them said they flew in for 2022. The sample is self-selected, and Arkia's CEO says flights are not full. Likud petitioned against the project; the Elections Committee rejected the petition after the organisers said they do not fund tickets. Nobody has measured how these voters will vote. The ballot-box data can answer three narrower questions.

What the surveys say

The Israel National Election Study (INES, Tel Aviv University) interviews the same people before and after each election. Before: "which list would you vote for?" After: "did you vote?". So for every list we can see what share of its intended voters say they voted. Five elections, 2019–2022, about 5,100 interviews.

Maps

One dot per ballot box, sized by eligible voters, at its polling station. The map follows the election picker. Scroll or pinch to zoom, drag to pan, hover for details.

The method, in one box

Caveats

• Hypothetical 2 hits its ceiling. In most elections one to four lists sit at 100% (Yesh Atid, Blue and White, Yamina, New Hope, National Unity). That means the model's assumptions are off: their boxes turn out highly for other reasons. Treat it as an upper bound on within-box gaps.
• Double envelopes are left out of the estimates (6–10% of votes) and added back unchanged in the seat counts.
• Map locations are approximate. There is only one polling-station layer (2026), so every election's boxes are matched to it by box number: 85% exact in 2022, 72–77% in 2009–2020. The rest go to the nearest-numbered box in the same town. The same matching gives the socio-economic cluster.
• The abroad correction uses the 2022 census for every year. Emigration patterns by town change slowly, but earlier years are an approximation.
• All ballot-box estimates use aggregate data (ecological inference). The survey is the only individual-level check, and it relies on self-reports.

Data and code