ENSLAVED BY POLLS

Public opinion surveys rule the media and the country's political class

By Robert Sarner, Canadian Jewish News, March 18, 1999

JERUSALEM - Few people this month are taking solace in the fact that we've now passed the halfway mark in a seemingly endless election campaign. Most Israelis would prefer to get the whole thing over with now and to spare themselves 10 more weeks of tedious politicking. But one group is delighted by this insufferably long electoral season. Since the Knesset voted in December to stage national elections on May 17, pollsters have been doing a bang-up business.

Even when politicians aren't running for office, Israeli pollsters are rarely idle. It's a toss-up who's more enraptured by public opinion surveys - politicians or the media since both commission them with increasing frequency.

During the current campaign, the country's two largest Hebrew dailies compete for the most newsworthy opinion poll results every Friday. To help boost sales of their fat, more expensive weekend editions, Yediot Ahronot and Ma'ariv publish survey results aimed at making headlines and capturing readers.

In recent weeks, Prime Minister Binyamin Netanyahu has been locked in a seesaw opinion poll battle with Labor Party leader Ehud Barak, his main challenger. Often on the same day, a Gallup Institute survey for Ma'ariv will show Netanyahu leading Barak by a percentage point or two while Yediot publishes a Dahaf poll indicating a Barak victory. Or vice-versa.

In Israel, like in many western countries, polls have become a multi-million dollar industry and an inescapable part of public life. Politicians and their advisers, almost enslaved by polls, have come to depend on them before making their next move.

Early this month, Attorney-General Elyakim Rubinstein ordered government ministers not to commission polls except for purely professional reasons until the elections.

"Ministers should not conduct opinion polls testing the satisfaction levels of the public or polls which include questions about elected officials or candidates," Rubinstein wrote in a special directive to cabinet members.

Polling is hardly an exact science but you'd never know it from the primacy given to opinion polls in Israeli politics and high-level planning. Despite their notorious unreliability, surveys are now a key part in government decision-making, having attained the status once reserved for top-secret intelligence reports.

"Public opinion surveys are playing an unprecedented important role in the management of affairs of state," Uzi Benziman wrote recently in Ha'aretz. "Not only do they gauge the public's mood and thereby serve as a legitimate assessment tool for both the national parties and the heads of political parties, they also determine what decisions are made."

Critics charge that surveys often distort reality with findings affected by several factors - the phrasing of questions posed to respondents; the expectation of those who commissioned the survey; the size of the sampling; and the manner in which the results are presented. The credibility of many polls is further strained by the fact that the main survey companies are hired by political parties and also serve the media in a potentially unholy alliance.

Some see polls in even more pernicious terms, insisting they have become a substitute for elections and undermine democracy by producing bandwagon or underdog effects. It's now a given of political science that published survey results affect voting behavior.

For their part, polling organizations say their work suffers from a public misunderstanding. They stress their independence and justify their polls as simply snapshots at a point in time of voter attitudes and intentions subject to change before voters reach the ballot box.

But how fair are polls? Pollsters do most of their interviewing over the phone so they rely heavily on phone directories for sampling purposes, thereby ignoring sectors of the population under-represented by such sources. As a result, the opinions of harder-to-reach groups such as soldiers, Arabs, haredim and new immigrants are not sufficiently taken into account, without mentioning others who refuse to be questioned.

There's something vaguely reassuring that for all their high-priced sampling, data collection and response analysis used in countless polls ahead of election day, these modern-day successors to soothsayers and crystal ball gazers often get it wrong. In 1996, pollsters predicted Shimon Peres would defeat Netanyahu for the premiership. In 1993, surveys indicated that incumbent Jerusalem mayor Teddy Kollek would beat his challenger Ehud Olmert in the capital's municipal elections. In 1992, then prime minister Yitzhak Shamir enjoyed unprecedented popularity with polls showing him easily vanquishing any candidate Labor could field. In all three cases, reality did not conform to predictions.

If recent elections are any indication, pollsters may well be eating humble pie for breakfast on May 18.