Poll: Kadima on 36 Seats, Labor on 18 and Likud on 14

Haaretz poll: 22 percent of respondents undecided, leaving 28 seats up for grabs; Yisrael Beiteinu at 7 seats.

A day and a half before polls open for the 17th Knesset, the latest Haaretz-Channel 10 poll puts Kadima on 36 seats, Labor on 18 and the Likud on 14.

The number of undecided voters has risen to 22 percent of those polled, meaning there are 28 Knesset seats still up for grabs.

The Labor Party, headed by Amir Peretz, has seen no change in the number of predicted seats since last week's poll. The Likud, however, has dropped one seat and Meretz-Yahad has picked up one, bringing it to six.

If the results hold true, it would make it easy for Kadima leader Ehud Olmert to form a stable coalition supportive of his plan for further withdrawals. A coalition made up of Kadima, Labor and Meretz would have 60 seats.

The right-wing bloc which includes the ultra-Orthodox factions amasses just 50 seats while Labor and Meretz manage to muster 32 seats.

Despite all the media attention and hoopla surrounding Yisrael Beiteinu and its chairman, Avigdor Lieberman, the poll predicts a loss of two seats to seven (from nine in the previous poll). The National Union-National Religious Party, in contrast, can claim credit for an improved showing in the latest poll with 12 seats (as opposed to nine in the previous poll).

The poll does not indicate significant change among the remaining factions - Shas would capture 11 seats, United Torah Judaism earns six seats, and the Arab parties total eight seats.

The Dialog company, under the supervision of Tel Aviv University's Professor Kamil Fuchs, commissioned the poll Sunday afternoon from a pool of 800 respondents representing a cross-sample of the public. The poll's margin of error stands at less than three percent.

This latest poll, however, was performed between 3 P.M. and 6:30 P.M. as opposed to prior polls whose samples were taken well into the evening hours, usually ending close to 8:30 P.M. In other words, this poll does not take into account the population group that arrives home after 6:30 P.M., a fact which is likely to impact the accuracy of the poll to a certain degree.