Is there going to be an Israel 2019b?

Today, the Knesset of Israel took the first step towards passing a bill to dissolve itself and set an early election, probably in early September. This came after Yisrael Beitenu leader Avigdor Liberman said he would not join an emerging government coalition that he claimed would be a halakha government. He was referring to the demands in such a coalition of the two Haredi Knesset factions, Shas and United Torah Judaism. (Halakha is Jewish law.)

It is still possible that this is all an elaborate ploy by PM Benjamin Netanyahu and his various allies to pressure Liberman into backing down. However, Liberman himself has said he welcomes new elections, and the preliminary reading of the elections bill passed with 65 votes, or exactly the number that the putative right/Haredi coalition would have if Yisrael Beitenu’s five seats are included.

In the 2019(a?) election in April, Netanyahu’s Likud won 35 seats, and the two Haredi lists 8 each. Add to those the seats of “soft right” Kulanu (4) and ultranationalist Union of Right Parties (5) and you get 60, exactly half the Knesset. These 60 seats could be sufficient to form and maintain a government, as long as Liberman and his YB do not vote against its investiture. Would they really vote with the left and Arab parties to stop its formation? It seems unlikely. Given the constructive vote of no confidence now in place in Israeli Basic Law, it would take at 61 seats to elect an alternative government in order to oust an already in-place minority government. It is even harder to imagine YB voting for any actual alternative at some future point in the life of the government. Nonetheless, Netanyahu clearly wants a majority coalition, even though pleasing all those small and essential partners would not make such a government notably more stable than a minority one.

It is also possible to imagine a flexible coalition deal in which YB is allowed to vote against the version of the Haredi conscription bill that Liberman objects to. However, the Knesset is under a Supreme Court deadline (oft-extended) to pass something to replace the current law, and it is not clear if any majority can be found for one. Yet the Haredi parties insist that the law be “fixed” to protect their constituencies from being drafted, because the reversion point is a law they dislike even more (i.e., there would suddenly be a requirement for many more enlistments from the ultra-orthodox constituencies, which is something that even the army is not exactly clamoring for).

So, unless a compromise can be struck by Wednesday, it seems the train is in motion for Israel to have two general elections in the same year for the first time in its history.

And I guess that means I’d be asked by the editors to update my chapter on the electoral system for the almost in-press Oxford Handbook of Israeli Politics and Society yet again.

Ukraine will have an early election

Well, it did not take long to learn the answer to my question. Yes, Ukraine will have an early election, as President Volodomyr Zelenskyy announced on 19 May in his inaugural address. And thus, no, the current electoral system will not be replaced just yet.

The election is expected to be in July, a scenario I already discussed in the earlier post.

In the context of all this, today (20 May) the Prime Minister, Volodymyr Groysman, announced his resignation.

There had been a report last week that a dissolution of the coalition in the assembly would prevent the new president from calling an early election (because it would buy the assembly time to attempt to find, under terms of the constitution, an alternative premier they can agree on). But evidently not.

Newfoundland and Labrador 2019

Now, this is a close election! Newfoundland and Labrador held its provincial election on 16 May. The result was the Liberal Party winning exactly half the seats–apparently. One of their losses is by five votes. Not five percentage points. Five votes. The apparent winner is the NDP candidate. There is likely to be a recount.

If the result holds, the NDP has won three seats in the 40-seat assembly, up from two in the previous election. The Liberals had 31 seats last time, so this is some electoral rebuke, even if it ends up back with a majority. The Progressive Conservative Party, which won 15 seats, has not conceded defeat, and the leader has said he will be calling on the NDP and two independents (who are ex-Liberals) to consider a possible bid for power if the Liberals do not win their 21st seat.

The Liberal vote total percentage was 43.9, PCs 42.6, NDP 6.3.

Ukraine: Possible early election and electoral reform (again?)

According to Hromadske (15 May), newly elected Ukrainian President Volodomyr Zelenskyy is considering dissolving parliament. Moreover, there is also consideration of electoral reform in a country that seems almost never to hold more than an election or two under the same rules.

Currently, an election to the national assembly is scheduled for 27 October. At about six months out from the presidential election, that timing is clearly a “honeymoon” election, and we know the impact those have. But waiting that long is not ideal for Zelenskyy, given he has almost no partisan support in the current assembly, having cobbled together his own campaign vehicle for his presidential run. His newly formed party is named after the TV show that made him famous, Servant of the People.

In the first round of the presidential election, on 31 March, Zelenskyy placed first with 30.2%, nearly doubling the runner-up, incumbent President Petro Poroshenko. In third place was perennial candidate Yulia Tymoshenko, with 13.4%. Four others had between 5% and 12%. There were thirty nine candidates in total! In the second round (21 April), Zelenskyy crushed Poroshenko, with nearly 75% of the valid votes cast. (I wonder if it would have been closer if Tymoshenko had made the runoff, or more competitive if ranked-choice voting had moved her or another candidate up into the final two.)

If the election goes ahead in late October as planned, it would be held with about 10% of the presidential inter-electoral period elapsed. The equation reported in Votes from Seats would imply that the Servant of the People party could expect around a third of the vote. (See my entry immediately after Macron’s win in France for the equation, graph of data from many countries, and discussion; note that our equation is based on first-round votes, and any fit to actual data would be much worse if runoff votes were used.)

One might understand why he thinks that is not enough. It is not clear to me what the date of the election might be if it is moved up. But let’s say it was 28 July instead. That would mean about 5% of the inter-electoral period elapsed, which would lead to an estimated vote share for the president’s party of… 34.5%. In other words, it is hardly worth the trouble!

Of course, the actual figure could be above these estimates–or below. One poll alluded to in Hromadske said that Servant of the People was the choice of only 25% of the people, a figure that would be a pretty disappointing honeymoon result. The more important point is that a three-month difference in timing does not really matter much for the honeymoon effect. Further, with no existing party to speak of, it might even be smart to allow more time to build the party and recruit candidates. And here is where the question of electoral system choice comes in.

Some electoral systems would be more demanding for candidate recruitment by a fledgling party than others. The current system is mixed-member majoritarian (MMM) and consists of 225 single-seat districts (plurality) and 225 closed-list seats (single nationwide district). Finding viable candidates to be personal representatives of the party in 225 districts is more of a challenge than filling out a closed list.

In fact, this challenge is mentioned in the Hromadske article, which states, correctly in my view, that “Returning to the closed list proportional electoral system would be most beneficial to the president-elect and his team.”  The “return” referred to here would be to the system used in 2006 and 2007 that used a single nationwide district for all 450 seats, and a closed list.

Other systems that are under consideration in the current parliament are a regionalization of the list component, and an “open list”. (I am not sure they really mean an open list, as that term has been applied misleadingly to a current local electoral system that is more along the lines of district-ordered list.) Regarding the latter option, Hromadske notes, “For Zelenskyy, it is easier to handpick a list of candidates [for a closed list], than to look for people, known locally in the regions, who could potentially win in an open list system.”

If the election is called early, of course the current system will prevail. Whether Zelenskyy can get the election earlier depended on precisely when he would be inaugurated (Hromadske explains). The date for the inauguration was just set in a vote on 16 May to take place on 20 May. This seems to allow time for an early dissolution, as the Hromadske article states that Zelenskyy and his team figured the last date for making such a decision was 27 May (though one loophole could allow that to be extended into June, perhaps).

Ukraine has developed a record of consistent changes of government and legislative majorities through elections, yet it has been anything but consistent with its electoral rules, or election timing.

Australia 2019

The Australian 2019 general election is 18 May. In fact, as I enter this text, it is only about an hour and half before polls open in the eastern part of the country (thus about 14.5 hours before they close in the west), even though it’s midday Friday where I am. So, for now I will leave the task of discussing election day and the early results to my several capable Australian commentators, as well as anyone else who wants to chime in.

I recommend this background piece in Inside Story on the history of Australian election results. Also this piece by Antony Green on calculating swings.

Spain 2019

Spain’s general election was on 28 April, the third one since December of 2015. For the third time in a row, the largest party will have under 40% of the seats. This time it is the Socialists (PSOE) in first place, with just 35.1% (123 of 350 seats) on only 28.7% of the nationwide vote. The last two times it had been the Popular Party (PP), but this time that party lost 69 seats to end up at 66, or 18.9% (on 16.7% of the votes). Two newer parties, whose breakthrough in 2015 had so much to do with the sharp decline in the two big parties’ votes and seats, are Ciudadanos (C’s) and Podemos. The C’s also gained, up 25 seats for a total of 57 (16.3% on 15.7% of the votes), while Podemos slipped considerably, down 24 seats to 42 (12.0% on 14.3% of the votes).

Then there’s Vox, the new nationalist party, which won 24 seats, which is 6.9%, on 10.3% of the vote. Note how significantly underrepresented Vox is, signifying its strength in rural areas which, under Spain’s electoral system, have low magnitudes (although with malapportionment, not as low as they would be if district magnitudes were redistributed to match current population shares).

The rest of the seats, as is typical, were mostly won by regional parties, with the Republican Left of Catalonia (ERC) being the largest, with 15 seats (4.3% on 3.9% of the total national vote). This is an increase of 6 seats over the last election.

Overall, the parties of the left did well. But the PSOE and Podemos remain short of a majority, unless they accept the support of the ERC. Which, of course, might only feed the strength of the right for next time, including Vox.

In some ways, it was actually a pretty typical Spanish result. The districted PR system, in which there are 52 districts for the 350 seats, once again allowed the two largest parties to be significantly over-represented (at least by the standards of the PR family of systems), and some smaller ones to be under-represented, except if they are regionally based. Of course, many of them are, and several of the smaller regional parties tend to be over-represented, due to having all their votes concentrated in a few districts (which often have moderate-low district magnitude). Of course, in the past–up to 2011 when the PP won a majority of seats–the two biggest parties were much more dominant in votes and therefore in seats than they have been in these past three elections. (The PP’s vote in 2011 was 44.6%; yes, manufactured majorities can happen under PR, such as when average district magnitude is on the low side.)

Note that, despite having had only 85 seats at the last election, the PSOE was the governing party going into this election, as a result of the constructive vote of no-confidence in the PP minority government in June, 2018.