We will never know for sure how much of a part polls played in shaping Prime Minister Benjamin Netanyahu's policies these past few months. We can only guess that it was a considerable one. His predecessors – Ariel Sharon, Ehud Olmert, and Ehud Barak – were "guilty" of the same thing. However, Israel's prime ministers were far less influenced by polling, simply because the political poll we know nowadays would have seemed like some futuristic fiction back then.
Follow Israel Hayom on Facebook and Twitter
Election polls also have a negative effect on how we, the voters, behave, even as we cast our ballots. Polls do more than portray a situation – they create it. It doesn't matter how many times pollsters like Mano Geva or Camil Fuchs emphasize that their results merely reflect a given situation. We, the public, have made polls into forecasts – a modern-day crystal ball.
In the modern era, fewer and fewer people are voting according to their values, interests, or worldviews. Many follow the polls and vote strategically. Some jump on the bandwagon of the predicted "winner." Others abandon parties whose values they identify with because polls give them a slim chance, or opt to bolster parties that appear as the underdog. Of course, there are also those who pick their parties based on its projected weight in a "bloc."
The parties themselves, and their leaders, allow the polls to drag them into this shallow discourse, and sometimes the math of the polls guides them, making positions and agendas marginal. Even the candidates for each list are sometimes chosen based on their electability alone, and the main question is – how many votes is each candidate "worth"? Theoretically -- and in practice – the "electable" candidate can be untalented and unsuccessful, but the moment a polls shows that he or she attracts votes – they will be given preference on the party list over people more capable but less electable than they are.
The surprise of 19
In recent years, polls have been the basis of party splits or partnerships. Recent notable examples include Blue and White and the decision by people such as Moshe Ya'alon, Ofer Shelach, or Ron Huldai to run as heads of their own parties, then drop out a short time later. The biggest weakness of polls is their short shelf life, which leads to short-term political thinking that leaves the long term out of the planning process. Decision makers now give ideology and worldviews a much more modest place, and in the world of the voters, ideas, plans, and views are given much less space, as well. Too much weight is given to who will win or lose and by how much, and where we, the voters, fall on the axis of their chances. In the political horse race that the media and the public have become addicted to, polls are the track, parties are the horses, and we are the cheering crowd, often without thinking.
This trips us up twofold: one time, when we abandon our true values because a poll leads us astray, and again when we do so based on a poll that is biased or incorrect and untrustworthy. Over time, polls have not been particularly accurate in predicting election results.
Subscribe to Israel Hayom's daily newsletter and never miss our top stories!
Only two years ago, Yishai Peleg ran an analysis of election polls spanning the years 1981-2015 for the weekly Makor Rishon. It turned out that the polls were many seats off from the real results of the elections in those years, getting an average of 22.7 seats wrong per campaign.
The most famous mistakes were about New Right in the campaign for the 21st Knesset, when not a single poll predicted that the party would fail to make it over the minimum electoral threshold of 3.25%. Polls also predicted that Isaac Herzog, who was head of the Labor Party in 2015, would beat the Likud by three seats. He lost by six. No poll predicted the success of Yesh Atid in 2013, when the party won 19 seats. In the March 2020 election, the last poll published by Channel 13 News predicted a tie between Blue and White and the Likud, but the Likud won three seats more. In the September 2019 election, most polls gave Likud the edge over Blue and White, but the opposite happened. There are many other examples.
We must cut off the polls earlier
Along with being held in thrall by the polls, the public, despite many reminders, refuses to look at their inherent weaknesses. Researchers such as Professor Gabi WEiman have been pointing out for years that the poll is an inaccurate compass, with entire sectors such as the Haredim, Arabs, new immigrants, and soldiers, going unrepresented. Many, sometimes 70%, of potential respondents aren't willing to answer, so they aren't represented either, and among the respondents, one-fourth to one-third say they haven't decided, that they are deliberating, or do not know for whom they will vote. Pollsters do not inform the public how these responses are weighted in the results. In other cases, respondents – for various reasons – lie to pollsters and intentionally dupe them. Top Israeli pollster Dr. Mina Zemach has suffered from this for years.
People who have followed polls over the year know that often, the questions are leading. Or as Nobel Prize laureate Ronald Coase, a British economist put it, "If you torture the data long enough, it will confess to anything." Often, the identity of the person who orders the poll influences its results. About 10 years ago, two polls were published days apart from one another. In the first, ordered by a "right-wing" official, only 41% of respondents expressed faith in the court system, while in the second, ordered by the universities of Beersheba and Haifa, 60% of respondents trusted the courts.
Recently, when the issue of an Israeli "annexation" was up for discussion, a poll by the movement Commanders for Israel's Security, which supports the two-state solution, showed that only 26% of respondents supported a partial annexation of Judea and Samaria. But another poll conducted by the Guttman Center at the Israel Democracy Institute (now renamed the Viterbi Family Center for Public Opinion and Policy Research) showed support reaching 50%. Large gaps were also seen in polls about the future of the Golan Heights, depending on who commissioned them. In general, the discrepancies can be explained by the wording of the questions, which to a large extent prompts different replies.
In the legendary TV series "Yes, Prime Minister," Permanent Secretary for the Department of Administrative Affairs, Sir Humphrey Appleby, once demonstrated how the public, through five "correct" questions, could be shown to oppose mandatory conscription. The goal of the exercise was to overshadow another poll in which 64% of respondents had answered "incorrect questions" and come out in support of the draft. Reality, it turns out, isn't far off.
In Spain, France, Belgium, and Italy, polls stop being published five, seven, 14, or 15 days prior to an election. In Israel, polls can run until three days before Election Day. The public and media's addiction to polls and the fact that polls don't only predict results, but create them, obligates us to decide on a longer poll-free period before we go to the polls.
It won't solve the problem, but it will make the public's vote a little cleaner and truer.
