Opinion polling for the 2026 Israeli legislative election concerns surveys of voting intentions ahead of the election to Israel’s Knesset. Such surveys typically ask which electoral list respondents would support if an election were held at the time of interviewing. Published results often emphasize projected parliamentary seats rather than vote percentages. These projections describe support under particular assumptions about participation and competing lists; they are not election results or guarantees of the eventual outcome.
Electoral framework
Israel elects the Knesset’s 120 members through nationwide proportional representation. Voters choose an electoral list, which may represent one political party or an alliance of parties. A list must receive at least 3.25% of valid votes to participate in the allocation of seats.
The threshold is particularly important when interpreting polls. A small change in estimated support can move a list above or below it, changing both that list’s projected representation and the distribution of seats among qualifying lists. Consequently, a change in a party’s seat projection need not correspond to an equally large change in its own measured support.
Israel’s parliamentary system also makes coalition arithmetic central to reporting. An absolute majority of the Knesset is 61 seats, a benchmark commonly used when discussing potential governing combinations. Nevertheless, a projected majority for a collection of parties does not establish that those parties would agree to form a coalition government.
What the polls measure
Voting-intention surveys and questions about preferred leadership measure different things. A respondent may prefer one person as prime minister while intending to vote for a list associated with another leader. Leadership-preference results therefore cannot be substituted for party-support estimates.
Surveys may also test hypothetical electoral arrangements, including new parties, joint lists, or alternative leaders. Such scenarios are conditional: a result obtained with one set of choices is not directly comparable with a result obtained using a substantially different ballot.
Likewise, published totals for a governing bloc or the opposition depend on which parties the publisher assigns to each group. Parties outside an incumbent coalition do not necessarily constitute a single alternative coalition.
Reporting and comparison
A meaningful record of polling distinguishes the polling organization from the media outlet or other body that commissioned or published the survey. Important information includes:
- Fieldwork dates, rather than publication date alone.
- Sample size and target population, including whether the sample covers the full electorate.
- Interviewing method, such as telephone interviews, online panels, or a combination.
- Treatment of undecided respondents, including whether they are excluded or assigned through a model.
- The electoral scenario tested, especially the composition of joint lists.
- Vote estimates and seat projections, where both are available.
Polls using different scenarios should be presented separately or clearly identified. Figures below the electoral threshold also matter, even when the corresponding lists receive no projected seats.
Uncertainty and limitations
Opinion polls are subject to sampling uncertainty, nonresponse, differences in population coverage, and assumptions about turnout. Weighting can improve a sample’s correspondence with known population characteristics but cannot ensure that every politically relevant difference has been corrected.
Published margins of error do not capture every source of uncertainty. In particular, a margin expressed in percentage points cannot simply be interpreted as a fixed margin in parliamentary seats. Threshold effects and the seat-allocation process make that conversion nonlinear.
Comparisons across several surveys can provide more context than a single result, but differences between polling organizations may reflect methodology as well as changes in public opinion. Polling averages also require care when their component surveys test different party configurations or cover overlapping fieldwork periods.