Three and a half months before election day, the polls are coming on an almost daily basis – and we let you track them with our Poll Trend tracker. Can you trust them, though? I must say you can't. Take a look at the following compilation of poll results from the previous round of elections and see why.
In the following table you can see what the polls said about four parties three and a half months before election day – compared to the final outcome. We gathered here a sample of several polls from October and early November of 2012, when the elections took place in late January of 2013. Note that much like what we see today, the early polls were taken when Israel's political map was not yet clear. In early October, the projected seats for Likud and Israel Beiteinu was higher, and the parties were still separate. At the end of the month they announced that they were running together and started losing seats. From their high of 42, they ended having just 31 seats. But we could not see this in early October 2012.
We also didn't know what Livni was going to do with herself until late in the game (until late November). She ended up having six seats – stealing away some of them from the Labor Party. Livni's late entrance is one of the reasons for which the projected number of seats for Labor in October and November 2012 – more than 20 in many polls – did not materialize on election day.
These numbers are a warning sign: three and a half months before election day, when the field is still unclear and the players and parties are still trying to figure out their way forward, the polls are not to be trusted as a projector of the true election results. They serve more as a guide to the parties on what they need to do, rather than as a guide to the observer of what to expect after the election. Look at the table (there's more following it):
Party/Poll
|
Maariv, Oct 11
|
Globes, Oct 19
|
Panels, Oct 26
|
Maariv, Oct. 29
|
Panels, Nov 2
|
Yediot, Nov 2
|
Outcome, Jan 22
|
Likud+Beiteinu
|
42
|
42
|
33
|
43
|
35
|
35
|
31
|
Yesh Atid
|
17
|
14
|
18
|
15
|
15
|
15
|
19
|
Labor
|
17
|
18
|
27
|
20
|
24
|
24
|
15
|
Habayit Hayehudi
|
7
|
9
|
13
|
8
|
11
|
7
|
12
|
The polls didn't accurately predict the number of seats for each party – but they were better, if not perfect, at painting the larger picture, that of the “blocs”. This is because people shift less often between blocs than they do between parties. A Likud voter could easily vote for Shas or Habayit Hayehudi, but voting for Labor or for Yesh Atid is more difficult for him (Kahlon could be a different story, because of his Likud background – but we'll have to see his list of running mates to better understand where he is headed. Lieberman could be a different story, as there are signs that he is ready to abandon the right-wing bloc).
Fact: two-three months before the last election the center-left could not claim a majority of projected seats, and it can't claim such a majority today. If votes don't move from the right-religious bloc to the anything-but-Bibi bloc, then Netanyahu will stay as Prime Minister.
Party/Poll
|
Maariv, Oct 11
|
Globes, Oct 19
|
Panels, Oct 26
|
Maariv, Oct. 29
|
Panels, Nov 2
|
Yediot, Nov 2
|
Outcome, Jan 22
|
Right-Religious
|
64
|
68
|
60
|
67
|
65
|
61
|
61
|
The last table has just two polls from days before election day – that is to show that even then there could still be great differences between the projection and the outcome. Israeli voters tend to weigh their many options until the very last minute.
Party/Poll
|
Haaretz, Jan 17
|
Yediot,
Jan 18
|
Outcome, Jan 22
|
Likud-Beiteinu
|
32
|
32
|
31
|
Yesh Atid
|
12
|
13
|
19
|
Labor
|
17
|
17
|
15
|
Habayit Hayehudi
|
14
|
12
|
12
|
And here are the latest numbers:
All right: Likud Beiteinu (Likud+Yisrael Beitenu)+Habayit Hayehudi+Strong Israel
All center: Yesh Atid+Hatnua+Kadima+ Kahlon
All left: Labor+Meretz+Hatnua
All religious: Jewih Home+Shas+Yahadut Hatorah+Am Shalem
All Arab: Hadash+Raam Taal+Balad