Elections 1: Where we are

Election time again. This post hopes to lay out the state of play before Tuesday’s general election in Israel. Tomorrow’s post will make a few predictions and things to watch out for.

The polls

First, the fundamentals. According to my own poll-of-polls using my own secret sauce averages*, this is the final position of the parties running in the election.

Yes it adds up to 121 seats. No, it doesn’t really matter. It’s an artifact of my process that I was too lazy to fix.

The party standings

Most polls give Blue and White (the mega-party formed from Benny Gantz’s Hosen L’Yisrael, Moshe Yaalon’s Talam Party, and the more-established Yesh Atid party of Yair Lapid) a slight edge over the ruling Likud. Sometimes this edge is as much as four seats, but in some polls it disappears altogether and they’re even or Likud is ahead instead.

Below them, things get messy. there’s pretty much only one medium-size party according to these numbers: Labor, on 10ish seats.

Every other party is polling at 7 seats or lower.

• Some are stable in the 6–7 range like Hadash-Taal, the Union of Rightwing Parties and United Torah Judaism.
• Some are more in the 5–6 range, like the New Right, Zehut and Meretz.
• Some are in the “5 and below” range which puts them at risk of dropping below the 4-seat threshold and disappearing entirely. These include Shas, Raam-Balad and Kulanu who are largely polling above the threshold
• A couple of parties that are polling below the threshold often: Yisrael Beiteinu and Gesher.

Note too that several of these small parties are actually joint lists of even smaller parties. Hadash-Taal and Raam-Balad are fairly obvious. The Union of Rightwing Parties is three smaller parties (Jewish Home, the National Union and the Kahanist Jewish Power party) and United Torah Judaism is really the Hassidic Agudat Yisrael and the Litvish Degel Hatorah).

This points to a very fragmented Knesset, especially when you remember that Blue and White is also a joint list.

The ‘blocs’

One of the biggest changes in Israeli politics in recent years is a much more significant polarization. It’s happened fast and it’s happened so totally that it erases the past with it.

Netanyahu’s 2009 government included the Labor Party, with Ehud Barak as his loyal Defence Minister. In the 2013 Netanyahu government, his first coalition deal was with Tzipi Livni, making her Justice Minister, in a government that also included several ministries for Yesh Atid and left the Haredi parties outside. Even at the beginning of his current term in 2015, Netanyahu spent months trying to entice Labor to join the coalition with the promise to make Herzog Foreign Minister.

None of this seems possible today. Politics in Israel has become all about the ‘blocs’, treated by discourse as immutable and immiscible.

Broadly, the ‘right bloc’ is those parties that will support Netanyahu and have publicly committed to recommend him to be Prime Minister. One of the most dramatic changes in the Israeli political landscape in the last decade is that the Haredi parties (UTJ and Shas) have joined the ‘right bloc’, promising to recommend Netanyahu as the next Prime Minister regardless of the outcome of the election.

In my model above, the ‘right bloc’, including Zehut (which is both ultra-nationalist and right-wing economically), has 64.5 seats on average.

There is no ‘left bloc’ as such. The centrist Blue and White, for example, includes former Likud members who are broadly to the Right of Kulanu as well as the economically-leftwing Histadrut Chairman Avi Nissenkorn. Many of the parties that aren’t promising to back Netanyahu have little in common. These parties alone cannot ever form a government, especially as the Arab parties don’t take part.

If Netanyahu gets his ‘right bloc’ of 61 seats or more, he will get the first chance to form a government. But don’t be too surprised if he chooses to reach out to Gantz first before looking rightwards.

If the ‘right bloc’ is below 61 seats, then things get interesting. Netanyahu is likely to still get first dibs at forming a government, but it’s possible that Gantz would get to try too.

A Gantz-led government would necessarily include parties from the ‘right bloc’, whether Kulanu or Likud itself or the Haredi parties. That could be difficult: the Haredi parties hate Yesh Atid with a passion.

The scandals

This has been a particularly nasty election cycle.

Benny Gantz has been accused of:

• having an affair
• Allowing Iran to steal sexually-explicit videos of him having an affair
• Supporting Iran
• Being mentally ill
• Being a paedophile
• doing secret deals with Arab ultra-nationalists
• Doing a corrupt deal with the Israel police

Most of this stuff is wild smears. But more credibly, Gantz has also been secretly recorded at private meetings, implying he could join a Netanyahu government and saying he doesn’t trust his co-leaders

On the other side, Netanyahu has had his own problems. He:

• was charged, subject to a hearing, with breach of trust, bribery and fraud in three different cases
• was accused of profiting off of a decision to buy unneeded submarines from Germany, an affair in which his advisers have been arrested
• was alleged to have illegally sold millions of shekels of stocks
• was challenged over his decision to allow Egypt to get advanced submarines, supposedly over the objections of the security echelon
• repeatedly and publicly pushed the Jewish Home party into a merger with the Kahanist Jewish Power (Otzma), a move which went against the party’s values and leader R. Rafi Peretz’s wishes.

Has any of this moved the needle? It’s hard to say. I suspect not. Before the Netanyahu charges, many supporters told pollsters that they wouldn’t vote for him if he was charged. But in reality, a couple of days later and there was no real impact on Likud’s poll numbers.

OK, that’s where we are. On Monday I hope to write a bit about the Known Unknowns that, even now, could upend our expectations.

(*the averages were determined by first averaging out the polls from a single polling company, then treating each company average as a single datapoint. This is to avoid double-counting any House Effects.)