By LARRY DERFNER

(February 19) -- To what extent do polls predict an election's outcome - and to what extent do they shape it?

Asked a couple of years ago about his chances of getting elected, a prominent Israeli politician started quoting public-opinion polls. This poll showed him increasing his lead, that poll showed him gaining strength with a crucial population sector, he said with enthusiasm.

'All that was off the record,' he hastened to say, then added: 'On the record, I don't pay much attention to public-opinion polls.'

There may be no greater, more frequently repeated lie in Israeli politics. Every politician of major stature spends a fortune on polls - especially during election season.

The clearest evidence of how much faith politicians place in public-opinion surveys was Yitzhak Mordechai's ascent to the leadership of the Center Party, which he achieved by barely defeating his rivals - Amnon Lipkin-Shahak, Dan Meridor and Ronni Milo - in polls taken by Dr. Mina Tzemah.

But faith in public-opinion surveys was shaken recently when the TV pollsters flubbed it badly on the Likud Knesset list.

Ma'agar Mohot, polling for Channel 2, picked Tzahi Hanegbi and Yisrael Katz to finish Nos. 1 and 2, and Geocartography, polling for Channel 1, picked Uzi Landau at the top.

In the end, Hanegbi and Katz were knocked down to 12th and 13th, and Landau finished ninth.

Pollsters point out that they do not predict the future, but only tell you the lie of the land at the time they took they poll. Baruch Mevorach, director of Ma'agar Mohot, and Dr. Avi Degani, director of Geocartography, both insisted that their poll findings, televised a week before the vote in the Likud Central Committee, actually caused many voters to change their minds.

'Members whose candidates finished high in the polls thought they didn't have to worry about them, so they voted instead for candidates who finished further down in the polls, in order to bring them up,' said Degani.

Also, he said, the polls showed Mizrahi candidates finishing relatively poorly, so party leaders spread the word to pump up the vote for Mizrahim - four of whom finished in the Likud's first five.

Finally, there were all the last-minute deals among candidates. No pollster could have tracked them, Degani added.

These and other difficulties put most leading pollsters off the Likud Knesset list. Neither Tzemah, nor Gallup, nor Hanoch and Rafi Smith ventured a survey on that vote.

Another surpassing difficulty was that a total of only 2,000 or people voted, so the polling sample was likely to be off, Tzemah noted.

But the pollsters - except for Ma'agar Mohot - redeemed themselves in the Labor Party primary, which was easier to read because there were over 100,000 voters. Deals were of little value.

Gallup correctly picked Shlomo Ben-Ami and Yossi Beilin as No. 1 and No. 2, and also got the next three - Matan Vilna'i, Avraham Burg and Uzi Baram - although not in that order.

Tzemah did a limited survey, asking respondents their top five favorites, and put Ben-Ami and Beilin first and second, missing out only by putting Haim Ramon third. Geocartography put Beilin, Ben-Ami and Burg in the top trio, with Vilna'i and Baram following.

Ma'agar Mohot, which took its poll a few days before the others, missed badly, putting Baram first, Binyamin Ben-Eliezer second (he finished seventh in the primary), Ben-Ami fourth and Beilin sixth.

When the poll was announced on Nissim Mishal's talk show, Elie Goldschmidt, who came out 20th in the survey and was sitting in the studio audience, was asked his reaction.

'I'm relieved,' he said, 'because if I did badly in the polls, I'll certainly do well in the election.' He was right: In the primary, Labor Party members voted him ninth.

While Degani insists that the publication of polls affects the election results, Tzemah is doubtful.

'Whatever changes occur probably cancel each other out. Some people might not vote for a candidate who does well in the polls because they think he's going to win anyway, but others may vote for him because they want to go with the winner,' she said.

Yitzhak (not his real name), a veteran Labor Party member from Jerusalem, said the polls had a marginal effect on his vote.

'I saw that Haim Ramon and Ophir Pines weren't doing so well in the polls, so I made sure to include them on my ballot,' he said.

Moshe (not his real name), a veteran Labor Party member from the Greater Tel Aviv area, said the polls had no effect whatsoever on his vote, because he didn't think they were accurate in the first place.

The question remaining is: Will the polls affect the outcome of the prime-ministerial race?

Surveys so far show that Labor leader Ehud Barak will finish well ahead of Mordechai in the first round and go into a runoff against Netanyahu, while the same surveys show that in a runoff, Mordechai runs much stronger than Barak against the incumbent.

Given the strong desire of veteran Labor Party members Yitzhak and Moshe to see Netanyahu defeated, if the polls on the eve of the first-round election, May 17, showed Mordechai with a chance of finishing ahead of Barak and getting into the runoffs, would they consider switching from Barak to Mordechai?

Yitzhak: 'No. I think Barak is the better candidate. He's smarter than Mordechai, and he's been clearer about his policies. Also, I was raised in the Labor movement, and this is my ideology.'

Moshe, who was also raised in the Labor movement: 'It's a possibility. I don't see it happening, because I don't see Mordechai improving his chances so much. But I don't want to close off any option.'

Tzemah said she didn't know if the polls on the eve of the 1996 election had affected the outcome. Like most other pollsters, she picked Shimon Peres by a slight margin -1 percent.

Although much of the public thinks the pollsters were embarrassed in the last election, Tzemah pointed out that Netanyahu had won by 1 percent, so her findings were only off by 2 percent - within the 3 to 4 percent margin of error.

'Very respectable results,' she concluded.

Links in this section:
More than a token victory?
The power game
Neighborhood watch
'I don't pay attention to polls'
Population shifts