The repeated failures by most pollsters to predict the results of recent elections should have kept reporters from contacting them or writing about weekly polls leading up to the Sept. 17 election. This is why this article is not about the number of seats that each party is predicted to win or how big the blocs will be. Instead, it focuses on the tools the pollsters work with to carry out the polls and process the results before making them public.
In the past decade, the limitations of research institutes have been made clear time after time. Only a small minority of the public contacted by pollsters are even willing to take part in surveys. In addition, the freehand the pollsters allow themselves in analyzing the responses of undecided voters is always problematic.
Follow Israel Hayom on Facebook and Twitter
We must also take into account the unwillingness of the Arab public and the haredi sector to answer pollsters' questions, as well as last-minute surprises that pollsters do not take into account. What's more, the law bans the publication of any polling results in the three days leading up to Election Day. All these contribute to the repeat failures we see on the part of pollsters to predict the final results. In the April 9 election, most erred by as many as 30-35 seats, which went to parties other than what they had originally predicted.
Israel Hayom reached out to a number of pollsters and asked them for insights into the process while focusing on their strong points – defining variables that will affect the final election results – rather than on predictions of the results themselves. Pollsters were asked about the projected voter turnout among supporters of each party and sector that comprise the major blocs; how certain respondents were that they would vote for the same party on Election Day that they picked in the polls; how pollsters handle the issue of undecided voters – those who intend to vote but aren't yet sure for whom; and what factors most affect Israeli voters' final decision.
Menachem Lazar, CEO of Panels Politics, who built his first election poll 31 years ago, thinks that Sept. 17 will see a lower voter turnout than the April election, when 68.5% of eligible voters cast ballots. He can already point to the one-fifth of his respondents who say they do not plan to vote, as well as another 10% who do not live in Israel. (Under Israeli law, private citizens living outside the country cannot cast absentee ballots.)
"This time, again, the number of people in the Arab sector who say they do not intend to vote is much higher than the number who say they do not intend to vote among the general population," Lazar says. In April, only 49.2% of eligible Arab voters cast ballots.
"In the last election, supporters of Blue and White were 5% less likely to vote than the general population of voters. On the other hand, the number of haredi voters who say they intend to vote is much higher than among the general public. If there is low voter turnout overall, the haredi parties will be the first to benefit, so their proportionate strength will increase," Lazar explains.
Lazar also looked at how sure voters were about their choices. He discovered that haredi voters were most certain for whom they would be voting, followed by Likud supporters; supporters of Yisrael Beytenu; Blue and White; Yamina; and the Democratic Union.
Lazar sees a serious problem for two parties when it comes to how certain voters are of their choice. The first is the Yamina (formerly the New Right) list. "On one hand, the nationalist-haredim have a hard time with the secular [Ayelet] Shaked as leader, and on the other, it's hard for the liberal religious to accept the dominance of [Bezalel Smotrich and [Rafi] Peretz."
The second such party – list, rather – is Labor-Gesher.
"In both cases, Labor and Yamina, the 'wavering' votes are wavering almost exclusively within the same blocs. Voters who say that they are voting Yamina now could wind up voting for the Likud, and voters who say right now that they'll vote for Labor-Gesher could eventually vote for Blue and White or the Democratic Union."
Professor Yitzhak Katz, CEO of the Maagar Mohot Institute, looked at another parameter that could indicate to what degree each party is holding onto its base from the previous election. He looked at how many supporters of each party planned to vote the same way and he, too, discovered that according to this parameter, Labor-Gesher is in trouble.
"Voter turnout will decide this election, because we're seeing virtually no movement between blocs. What will be most important is not who votes, but who doesn't."
"Only 27% of the people who voted for Labor last time say that they will vote for Labor this time. On the other hand, 32% of those who voted for Labor last time plan to vote for the Democratic Union this time, and 14% plan to vote for Blue and White. This definitely explains why Labor-Gesher is bobbing around near the minimum electoral threshold," Katz says.
However, according to Katz's findings, 70% of people who voted for Blue and White in April plan to vote for the party again, while 2% will be moving to Labor-Gesher, 3% to the Joint Arab List, 4% to Yisrael Beytenu, 4% to the Democratic Union, and 2% to Yamina.
What does this parameter show us when it comes to Yamina and the Likud?
"[Some] 65% of Likudniks will vote for the party again [on Sept. 17], 7% will move over to [Avigdor] Lieberman, 5% to Yamina, 4% to Blue and White, and 17% still aren't sure," Katz says.
"For Yamina – a list that includes two parties that ran separately in the last election – the situation is more complicated: 64% of people who voted for the New Right say they will vote for Yamina, but only 40% of those who voted for the United Right [Habayit Hayehudi and National Union alliance] say they will vote for Yamina."
Like Lazar, Katz also looked at many people who voted in April would be voting on Sept. 17, if not necessarily for the same party. He says that the results provide an indication of what kind of voter turnout can be expected for the different blocs. Both Katz and Lazar predict nearly zero movement from one bloc to another.
Katz found that 82% of Labor voters from April plan to vote on Sept. 17, compared to 74% of supporters of the Democratic Union and United Torah Judaism; 69% of Blue and White supporters; 68% of Yisrael Beytenu voters; 65% of Likud voters; 62% of Yamina voters; and 55% of supporters of the Arab parties.
Rafi Smith, head of Smith Consulting, offers interesting information about the source of Lieberman's strength. Despite the Likud's attempts to make him go away, his support in the polls has doubled from five seats in April to 10 seats or more in recent polls.
"Lieberman is getting about two seats from the Likud, two from Blue and White, and some from Yamina … and Kulanu," Smith explains.
Like his colleagues, Smith foresees a difficult election for Labor: "Polls show it losing two seats to the Democratic Union. That means they have a very low rate of die-hard support."
Smith doesn't provide numbers about the wavering votes for the two biggest parties – Likud and Blue and White – but he says that generally speaking, Blue and White supporters tend to be less definite.
"If there isn't any major event in the south, the reality there will have little influence on the election."
"What will determine the results is voter turnout. If all Israelis were to turn out and vote, it would be good for the Right. Demographically, the public has been trending right-ward for years. Low voter turnout would be bad for the Right. A 5% increase in voter turnout in the Arab sector would be enough to give the Left another seat," Smith says.
Smith has also found that supporters of Gesher, which joined up with Labor, and Kulanu (which joined the Likud), could split their votes among several parties.
"The profile of Gesher and Kulanu voters isn't one-dimensional, so there is no one party they are flocking to. We can find them in both blocs, in this party or in that party. In comparison, people who voted for the New Right or the United Right [in April], who found a joint home in Yamina, are more homogenous – certainly closer to each other than Kulanu voters are to the Likud, or Gesher voters are to Labor," he says.
Smith is pessimistic about the prospects for the far-right Otzma Yehudit, which as of this week is still in the race.
"I think they will get about 70,000 votes. To pass the minimum threshold, you need 130,000-140,000 votes. That's an insane number of people, a little less than the population of Ramat Gan and much higher than the number of residents of Raanana or Kfar Saba. It's a daunting number. At least based on what the polls are telling us right now, it's hard to believe they'll make it."
Aside from voter turn-out and pure numbers, what can the pollsters tell us about the concerns of the undecided voters – not to mention Israeli voters in general? What issues will determine their final vote?
Professor Camil Fuchs, the veteran pollster for the Dialogue Institute, discussed the subject last week in a lecture at the Israeli Institute for Democracy. In his presentation, he revealed a new study conducted in July and August that collated the results of six polls conducted among 2,173 Jewish Israelis.
Fuchs did not ask which parties they intended to vote for. He asked which blocs they would be supporting: Right, center-left, religious-haredi, or Yisrael Beytenu. The results indicate that among Jewish voters, the right-wing/religious/haredi bloc is in a dead heat with the center-left/Yisrael Beytenu, with each winning 42% of the public's support. The Arabs could tip the scales, but Fuchs' survey did not include that sector, and in light of the low expected voter turnout, it's doubtful that will happen.
Fuchs found that the older Jewish Israelis were, the more likely they were to vote for the center-left.
"Young people lean much more to the Right. The two factors that are the best predictors are previous votes and how religious the voters are. In general, secular Israelis vote more for the Left and the religious vote more for the Right. Among secular Israelis, there is a difference between immigrants from the former Soviet Union, who lean much more toward the Right and Yisrael Beytenu, and the rest, who lean toward the Center and the Left."
Fuchs also analyzed voting according to location and, not surprisingly, found that support for the Right is stronger in Judea and Samaria, in southern Israel, and in Jerusalem. In central Israel and the greater Tel Aviv area, support for the Left is much higher.
Smith adds that "In places with high voter turnout by the religious and a high percentage of Mizrahi voters, you'll always have more right-wing votes. It doesn't matter if it's in the north, the center, or the south. On the other hand, secular Ashkenazi voters will vote less for the Right and more for the center-left.
"Likud voters say that security and defense issues clinch their votes. Voters on the Left say that social issues are what decide, but what really affects recent elections is the person who heads the party. In the past few elections and this coming one, the question was and is: yes or no to Bibi?"
Smith, Katz, and Lazar agree that at least for now, events on the Gaza border and the western Negev don't have too much of an effect on voters, who have apparently gotten used to the situation.
"Gaza has turned into a chronic problem," Lazar says. "So it has less of an effect on voters."
Katz thinks that "If there isn't any major event in the south, the reality there will have little influence on the election."
"In this election, like the last one, there's a strong personal element," Lazar says.
"Other than Netanyahu, the polls indicate that there are two personalities for whom people are voting, rather than their agendas: Ayelet Shaked in Yamina and Gabi Ashkenazi in Blue and White," Lazar adds.
Right now, the undecided or unsure comprise 15% to 25% of voters.
According to Fuchs, "Voter turnout will decide this election because we're seeing virtually no movement between blocs. What will be most important is not who votes, but who doesn't."
