With less than a week to go in the American Presidential (and other levels) election campaign, it is useful to look at Israel’s law regarding the publication of polls during the last five days of an election: prohibited! The U.S., of course, has no such law (free speech über alles).
Israel is not the only country with a poll publication prohibition. In fact, approximately two-thirds (!) of all countries in the world with elections have some sort of pre-voting day restriction on publishing polls. Some countries – e.g. Bolivia, Cameroon, Honduras and Tunisia – embargo polls for 30 days or more before voting day, while another ten (some highly democratic, e.g. Chile and Italy) have poll blackouts for at least two weeks pre-election day (https://www.researchgate.net/publication/330509451_Freedom_to_Conduct_Opinion_Polls_A_2017_Worldwide_Update)
All of this leads to the question: why? Why would otherwise perfectly functioning democracies restrict “polling speech” during the latter part of an election campaign? The basic answer is that there is a general feeling that the voters might be unduly influenced by such survey results, i.e. many people will vote by following the pack and not for reasons of real personal preference. As a political scientist, I can say two (somewhat contradictory) things: that’s probably true to a limited extent; there is virtually no hard evidence for that happening.
However, there’s a completely different problem with polling publication (during the entire campaign) and it has to do with the news and what is called the “media agenda”. There are actually three types of agenda: media (what the news purveyors focus on), political (what the politicians want you to think about), and public (what the citizenry actually cares about). Each influences the other, but in our hyper-media age (traditional media, e-media, and social media), it is the media agenda that rules supreme. And that’s the source of our problem.
During elections campaigns there are three types of news: policy, personality, and polls. The latter is actually called “horserace journalism”: who’s ahead, who’s behind, and who’s fading or coming on fast? In and of itself, that sort of thing interests all of us, but if it overwhelms the other two, then democracy is in trouble.
It is no coincidence that over the past several decades, as polling became more widespread and sophisticated (not merely gross numbers but group or regional breakdowns) horserace journalism pushed policy discussions off the public agenda. In Israel, with its traditional party list system, polling has also strengthened “personality” issues as a driving force in the campaign, with the pollsters starting to ask questions like “Who is most fit to be prime minister?” – this, even though there are no direct elections for that post!
None of this is to suggest that polling is useless. Paradoxically, it is most helpful in between election campaigns (“normal politics”) in order to clue government leaders as to what the public wants on specific issues. This can even be the case when a survey does not ask about a specific policy question. For instance, it is clear that the Likud’s fast decline in the recent polls is in large part due to Netanyahu’s failed mid-summer Corona policy – not only because he didn’t always listen to the medical professionals but also more recently for not holding the Haredi (and in the first wave, the Israeli-Arab) sector to account for not adhering to the government’s Corona regulations.
Returning to campaign polling: looking back at this U.S. election cycle, one would have to search with a magnifying glass to find any serious news coverage of policies. It’s been almost all “Biden maintains a big lead” (the horse race) or “Trump is angry at his Cabinet” (personality). Of course, a candidate’s temperament is important (President Trump is a good/bad example of that), but in the final analysis we want to know what our leaders plan to do if and when they enter office. The candidates’ policies could be poles apart, but by unduly focusing instead on the polls we are hardly able to know if that’s indeed the case.