The localities that predict the national result
Some localities' results move almost exactly as the national result does, election after election. They do not represent the country — they swing like it, and that is what makes them useful on election night.
Who moves like the country
The ranking is by swing: how close the change in each list in the locality, from one election to the next, is to that list's national change. It is measured in percentage points, and lower is better. The second column is a different measure altogether — how much the locality looks like the national result — and a locality can excel at one and fail at the other.
| # | Locality | Swing against the national | Likeness to the national | Elections | |
|---|---|---|---|---|---|
| 1 | Rehovot | 0.44 pts | 0.73 (1) | 10 | The results → |
| 2 | Haifa | 0.46 pts | 0.93 (3) | 10 | The results → |
| 3 | Netanya | 0.53 pts | 1.14 (5) | 10 | The results → |
| 4 | Petah Tiqwa | 0.53 pts | 0.88 (2) | 10 | The results → |
| 5 | Yavne | 0.56 pts | 1.21 (10) | 10 | The results → |
| 6 | Hadera | 0.59 pts | 1.14 (4) | 10 | The results → |
| 7 | Holon | 0.60 pts | 1.20 (9) | 10 | The results → |
| 8 | Mevasseret Ziyyon | 0.60 pts | 1.28 (14) | 10 | The results → |
| 9 | Azor | 0.61 pts | 1.19 (8) | 10 | The results → |
| 10 | Qiryat Atta | 0.61 pts | 1.30 (18) | 10 | The results → |
| 11 | Pardes Hanna-Karkur | 0.62 pts | 1.15 (7) | 10 | The results → |
| 12 | Nahariyya | 0.64 pts | 1.30 (17) | 10 | The results → |
| 13 | Bat Yam | 0.64 pts | 1.37 (22) | 10 | The results → |
| 14 | Gan Yavne | 0.65 pts | 1.22 (11) | 10 | The results → |
| 15 | Gedera | 0.65 pts | 1.23 (12) | 10 | The results → |
And who does not
The other end of the same query
The localities that move least with the national result are not bad predictors — they point to a different question. It is exactly what the first axis of the political map measures, and it turned up here again without being looked for: two separate analyses, the same conclusion.
| Deir Al-Asad | 3.20 pts | 8 |
| Daburiyya | 3.16 pts | 8 |
| Sakhnin | 3.11 pts | 10 |
| Umm al-Fahm | 3.10 pts | 10 |
| Ma'ale Iron | 3.07 pts | 9 |
How it is measured
For each locality, and each list that stood, the list's share in the locality and its national share are calculated — in the same election. Swing is the mean of |local change − national change| between each two consecutive elections; likeness is the mean of |local − national| itself.
The ranking includes 135 localities — those with at least 3,000 valid votes counted in at least 8 of 10 elections. The bar is not a judgement on small localities but arithmetic: in a locality of 500 votes, one extended family moves a list by a percentage point, and the "deviation from the national" would measure the size of a family, not politics.
What this is not
A bellwether locality is not a prophecy. It is a description of what happened in ten elections already counted, and no rule obliges it to carry on. It is useful on election night exactly as far as the past carries on — no further.