Trends and analysis

Turnout: the current behind the cleavage

A locality's position on the map is derived from the make-up of its vote, not from how many came to the polls, so turnout is a separate story, and an illuminating one. In Jewish localities it barely moves over twenty years, between 62% and 69%. In Arab localities it swings by 20 points: from 45% in 2021 to 65% in 2020. And that peak fell in exactly the election in which the cleavage was narrowest: when the quieter camp speaks at full voice, the distance between the camps shrinks.

An estimate, not a measurement Turnout is the number of voters divided by the number eligible to vote, over the same localities as the cleavage chart. For the 2006 election the source data has no count of eligible voters, so the line breaks there; it is not filled in by guesswork.

The cleavage that narrowed, and opened again

The map's first axis is the Arab/Jewish cleavage. Following each camp's centre of gravity on that axis, election by election, gives a measure of how similarly the two electorates voted, and it is not a line that steadily falls or rises. One might expect a cleavage only to deepen; the data says otherwise. The distance between the camps narrowed from 15.4 in 2003 to a low of 9.1 in 2020, and then opened again to 14.6 in 2022. This is not an effect of weighting: with every locality counted once, whatever its number of votes, the same narrowing appears, and more sharply, from 20.4 to 9.9. The narrowing and the reopening move together with turnout among Arab voters (see above): the cleavage is narrowest when the Arab vote is at its peak, and opens when it drops. In the 2022 election an Arab list (ד) even fell below the threshold.

An estimate, not a measurement The distance is measured in the map's units, so what means something is the trend, not a single value. And it describes places: how similarly the two kinds of locality voted, not people moving between them.

The localities that moved most

Measuring change on the map, rather than in raw vote shares, solves a problem: a party that only changed its name sits at the same point, so a renaming counts as no movement. These are the localities that travelled furthest between the first election and the last, limited to localities large enough (8,000+ votes) for the movement to be a real shift and not the noise of a small electorate. Each movement is split along the map's two axes: on the cleavage axis, converging or diverging between the camps, and on the map's second axis, the movement within the camp.

The external envelopes: the vote that grew

External envelopes, the votes of soldiers, diplomats, hospital patients and prisoners, cast away from the locality where they live, do not enter the map's calculation, but each election can be projected as a separate point on the same axes. The dark trail on the map above follows them over the years.

Two things stand out. The volume nearly doubled: their share of all votes rose from 5.0% in 2003 to 9.6% in 2022, with the sharp jump in the elections of the coronavirus period. The position: on the map above, the trail sits over the years on the side of יהדות התורה · כח להשפיע בהכוונת רבי · חזית יהודית לאומית.

The data shows growth, not a failure The rapid growth stirred public debate and oversight. Shown here is only what is in the Elections Committee's data, the volume and the position, and no claim at all about the soundness of the vote.