top of page

עליתי לכם
עד כאן

MethodologyAbout

How we build this guide


This page explains where our information comes from, how we tell facts from analysis and estimates, and how each figure on the site is calculated.


Who is covered

The lists running in the 2026 Knesset election that are in our research scope, and the people on their 2026 slates. Only these lists and people have pages. Other parties and people (for example, a Knesset faction that is not running in 2026) are named as plain text where they appear in the record.


Where our information comes from

We use the official record first:

  1. Party platforms and official channels.
  2. The Knesset record — plenum votes, bills, committee protocols, committee membership, parliamentary questions — and the Knesset lobbyist register.
  3. The party registrar, the Central Elections Committee, the State Comptroller's campaign-finance filings, the Competition Authority's list of concentration entities, and published government procurement awards.
  4. Statements by a party's leader or senior representatives, quoted and dated.
  5. Reputable news reporting, only to fill gaps and always attributed.


Encyclopaedias, profile pages on social networks and other voter guides are used only as pointers to an original source — never as the source we publish. Each source we publish is captured and kept as evidence, with an independent archive copy where possible.


Stated positions and voting-record positions

  • Stated position — what a list says on a proposition, from its platform or an official statement, with source and date. An older statement stays current unless the list has revoked or replaced it; its date is always shown.
  • Voting-record position — our analysis of how a faction voted in the Knesset on bills linked to a proposition. Each bill is linked to a proposition by a person, never automatically, and a voting-record position is shown only when it rests on at least 5 counted votes on at least 2 different bills or other legislative items. It describes a voting pattern, not the party's self-stated position. Where the faction was in government at the time, coalition discipline can shape a vote; we account for it and say so beside the position.


We show the two side by side and never blend them into one score.


Words and deeds (contradictions)

A contradiction case shows exactly two facts: the list says X (its stated position, with source and date) / its candidate voted Y on [date] (the Knesset record). We show the faction the candidate voted with at the time. Abstaining, not voting or missing information is never treated as a contradiction. We do not add a motive or a label of our own.


K25 participation

For a person who served in the 25th Knesset: the votes they cast (for, against or abstain) out of the plenum votes held while they were a member, from the official record. "Did not vote" means a vote they were eligible for with no recorded vote. A candidate who was not a K25 member is "not applicable", never zero. For a list, we add up the opportunities and votes of the people on its slate now, and show how many people that covers.


Slate composition

  • Women — counted from how a source refers to each candidate. Where no source states it, we may infer it (for example from the given name) and mark it "inferred"; unknown stays unknown. Shown for the top 10 and top 20 of each slate.
  • Born abroad — the birth country as a published source states it, for the whole slate; unknown stays unknown. It describes a place of birth only, nothing about citizenship, language or immigration.
  • Age — on election day, 27 October 2026, where a birth date or year is published.


On the Data and charts page, a list's figures count the people on that list now, including members who served in another faction before.


Lobby positions and agreement rates

  • A lobby position is an interest group's recorded support or opposition on a bill, from a Knesset committee protocol or the group's own published statement.
  • An agreement rate counts, on the bills where a group recorded support or opposition, how often a faction's majority (or a candidate's own vote) went the same way as the group and how often it went the other way. We show both directions, the number of bills compared, the contested bills (where both sides had at least 10 votes in the plenum) and the date range, and every bill behind the figure. A rate is shown only when at least 5 bills were compared; below that we show the individual bills only. Abstentions are counted separately.
  • An agreement rate is a fact about the record. It is not evidence of influence, and we never say why anyone voted as they did.
  • On each lobby group's page, a chart shows the candidates who most often voted as the group argued and those who most often voted against its position, with the number of bills for each. When 10 or fewer candidates qualify, the chart shows all of them.


The committee-weighting estimate (our estimate)

On some competition-related votes we mark whether the candidate sat on the Finance, Economics or a competition committee on the day of the vote. We treat such a vote as a "stronger" signal than a vote by a member outside those committees. The committee seats and the vote are facts; the weighting is our estimate and is always labelled so.


Campaign finance

From the State Comptroller's filings: donations, loans and loan guarantees, to parties and to candidates in primaries. Names, residence and employer appear as filed. Donations, loans and guarantees are different things and are never added together — a guarantee backs a loan; it is not money received. A role that a backer holds at a company, or that a candidate holds at an organisation, is shown as a documented fact; it is not evidence of influence, benefit or wrongdoing.


Market concentration and procurement

Market-concentration figures count control chains and firms on the Competition Authority's list of concentration entities, by industry; counts below 5 are shown as "fewer than 5". Procurement totals add up published government awards to organisations in our register; awards we could not link to an organisation are left out, and we say so. Industries are our own classification of each organisation and are labelled as such.


Find a party

Your answers are compared only with lists' stated positions on the same propositions; voting records are shown for context but not scored. A list is scored only when at least 5 of your answers can be compared with its positions. The result is your personal comparison, not our recommendation. Your answers stay in your browser.


Ask i-polit

Ask i-polit answers only from the published information on this site, using fixed answer templates with a source and date on every fact. An AI model is used only to choose which template answers your question; the model's own text is never shown, and it never writes statements about parties or people.


Languages

The site works in Hebrew, English, Russian, Arabic, Ukrainian, French, Spanish and Amharic. Names, official bodies and labels are shown in every language from a reviewed glossary; a transliterated name is marked as such. Position texts exist in Hebrew, English, Russian and Arabic; in other languages we show one of these, labelled with its language.


What we don't do

We do not endorse or oppose any party or candidate, appeal for a vote, rank parties as better or worse, assert motives, or publish allegations we cannot source to our standard. When we cannot verify something, we leave it out.


Corrections

If something here is wrong, out of date or misattributed, please tell us — see Corrections.

How we build this guide

bottom of page