Israel Elections: Are the voter polls even accurate?

ELECTION AFFAIRS: Israelis love polls. But are they accurate, and do they help or harm our democracy?

THE RESULTS of the exit polls on channels 11, 12 and 13 are shown on large screens at Blue and White Party headquarters following the September 2019 election. (photo credit: AMIR COHEN/REUTERS)
THE RESULTS of the exit polls on channels 11, 12 and 13 are shown on large screens at Blue and White Party headquarters following the September 2019 election.
(photo credit: AMIR COHEN/REUTERS)
 It’s a month before the election, and the land is awash in polls.
Everywhere you turn there are polls: Channel 11, Channel 12, Channel 13, Channel 20, Maariv/The Jerusalem Post, Israel Hayom, Walla, Radio 103fm.
And not only in the media. On your phone, in your SMS file. The polls are everywhere; you can’t get away from them.
But are they good? Meaning, are they accurate, fair and beneficial to the democratic process? The answers to those questions lie very much in the eye of the beholder.
Udi Lebel, a professor specializing in political psychology at Bar-Ilan University’s School of Communications, has for the last two years been studying so-called push polls – those surveys that constantly land on your phone with one or more questions. These polls, Lebel said, try to shape reality, not reflect it.
These surveys are more political marketing tools than actual polls. In fact, very rarely is anyone sitting on the other end of the phone tallying up the results to the survey questions. Rather, the purpose of these questions is to plant ideas in your mind to impact how you vote.
The first function of these polls, said Lebel, is to disseminate information that would be difficult to disseminate through traditional means because the information is, well, sketchy.
To illustrate this technique, Lebel said that two campaign cycles ago, a push poll circulated asking respondents whether they would vote differently if they knew that the wife of Yamina head Naftali Bennett did not lead a religious lifestyle.
In this case, Lebel said, the survey was nothing more than a tool to circulate a rumor that – because of journalistic standards – would be difficult to get into print. But by using the information in the form of a question in a survey, the initiator of the survey got the idea out there without having to get Bennett’s or anyone else’s response. No one cares at all about how the respondents answer the question, and there is no statistical company crunching the numbers. In this case, the intent was solely to smear.
This technique, Lebel said, paradoxically works both on people who have decided to cut themselves off from the news, as well as those inundated by it.
Those who may have decided to stop reading papers have, Lebel said, certainly not cut themselves off from their cellphones, and will see the surveys and their leading questions. He said that focus groups he has conducted demonstrate that people bombarded by news don’t necessarily remember what they heard, but will have a better chance of remembering a piece of information if it lands in their hand in the form of a survey on their phone.
People, ironically, also have a greater tendency to believe a survey they have been sent, because the whole idea of polls is wrapped in an aura of being objective and scientific.
“If you get a survey from Labor or the Likud, you will be skeptical, but from a polling institute? The assumption is that it is scientific, not something manipulative trying to plant ideas into your subconscious.”
Which, he asserted, is precisely what many of these push polls try to do.
“A second function of these surveys is not information injection,” Lebel said, “but cognitive framing. This means that you put into people’s consciousness something you may not want to say outright.”
A real-time example of this, he said, was a question on a recent push poll that asked, “If you knew after the election that there will be a merger between the Likud and Yamina, how would you vote?”
Clearly, Lebel said, this was not a Likud or Yamina poll, but rather, a poll circulated by an opposing party that wanted to plant the idea that there is a real likelihood that Yamina would merge with the Likud after the election, so if you don’t want Prime Minister Benjamin Netanyahu, it would be best to stay away from Bennett as well.
“This is meant to introduce a cognitive script in your mind, so the reader thinks it is a real option,” he said.
Push polls also advance agendas through the order in which they list the possible answers. For instance, a survey may ask, “Who do you think is best suited as prime minister?” and then list Netanyahu first, Yesh Atid’s Yair Lapid second, and New Hope’s Gideon Sa’ar or Bennett way down the list.
This creates in the person’s mind the idea that the race is really between Netanyahu and Lapid, and that the other candidates are just afterthoughts.
That is all regarding push polls, but how about the effectiveness of the television and newspaper polls that people see constantly?
Here Lebel said that the public does not realize the degree to which it is being manipulated, as there is a billion-dollar industry behind these polls, involving scores of subcontractors and marketing people who have different financial and political interests. The viewer, Lebel said, does not know of the different competing interests or even how the polls were carried out.
REGARDING THE polling industry in the country, Jerusalem-based pollster Mitchell Barak said that there is a “built-in conflict of interest among those doing polls, because there are not enough companies conducting polls in the country.” For instance, a pollster for one of the main media outlets could at the same time be conducting internal party polls.
The average citizen sitting at home watching the polls on the evening news needs to realize, Barak said, that what they are seeing are trends.
“Israelis don’t really make their final decision until right before the election – from midnight Friday to [Election Day] Tuesday, because you never know what is going to be the main issues on the day of the election. Coronavirus? Terror? Peace with Saudi Arabia? Iran? You don’t know what the issues are, especially in a place like Israel.”
As a result, he asserted, “a lot of people kind of change their minds at the last minute when they get the last Friday newspaper before the election, [and] when they go to their family’s house, then they sit down and really focus on what’s happening.”
Mina Tzemach, long one Israel’s most high-profile posters, famously said – after her Channel 12 exit poll in April 2019 was very wide of the mark – that people lie to pollsters.
Barak said he does not believe that is a widespread phenomenon.
“People give the answer that they think at the time, but here things are very fluid, and the situation changes all the time,” he said.
One problem with the major media polls, Barak said, is that they are skewed against the smaller parties.
“Polls are reporting on whether parties will get past the 3.25% electoral threshold, but that is a problem because that is less than a poll’s margin of error,” he said. “It is not responsible for a media outlet to say that a party is not getting over the threshold.”
Polls that show one party or another not making it over the threshold, he said, very much shape how people vote, as people are less likely to vote for a party that does not seem to have – according to the polls – a chance of making it into the Knesset.
But Shmuel Rosner, a columnist and editor of a new Israeli polling website, – patterned after the RealClearPolitics polling aggregator and political news website in the US – took issue with Barak’s complaint, saying the margin of error in polls is for the entire poll, not just one party.
Rosner said that it is important to look at a number of polls, and if a smaller party is not getting into the Knesset in a number of them, that is a sign that the party is in “a danger zone.” Informing the voter of this, he said, is important information, akin to letting voters know if the leader of a major party is dying, because that, too, would likely affect how people might vote.
The media polls are “more accurate than most people think, and less than most would want,” he said. But, he pointed out, in a heterogeneous society with a multiparty system where 1,000 votes one way or another could cause a party to fall below the electoral threshold and change everything, absolute precision is unrealistic.
“There is a need to look at polls as you look at newspapers and television – they provide information,” Rosner said. “The information is helpful to politicians who look at it to decide how to act, and to voters who can use it to decide for whom to vote.
“Imagine a world without polls,” Rosner said. “You have no idea whether the Likud will get 30 seats and Meretz four, or vice versa. You want that information. It may be imperfect, but it is close to reality, and it is vital.”
Rosner said that the precision of polls over the last decade, though not maximal, has not been “that bad.”
“Take the last three elections,” he said. “We knew through the polls that the elections would be close, and it could depend on one or two seats falling this way or that. The results were not dramatically different.”
As to why there is a popular perception that the polls always get it wrong in this country, Rosner said that this is because the public is looking at the final results.
“If the polls predict a party’s result and are off by only one seat, that is pretty good,” he said. “But if that one seat is the difference when it comes to forming a coalition, then it is pretty bad.”