In addition to Israel, there is also a general election today in Denmark. I don’t really know much about Danish politics, but I hope readers who are more informed might have comments as results become known.
One particularly interesting feature of the campaign that I will note is that former PM Lars Løkke Rasmussen is leading a Moderate (with a capital M) Party in this election that is campaigning outside the left and right blocs. If neither bloc wins a combined majority of seats, and his party does particularly well, it might attempt to lure the relatively more moderate (with a small m) parties from each bloc to split and form a center-spanning coalition. Otherwise, he could just be the “kingmaker“–a term I dislike, but which fits better than usual in a case of a party that has campaigned with indifference towards left or right and could strike a coalition deal with either. (That second link is actually about the first scenario I mentioned, and not the real “kingmaker” scenario.)
The incumbent government is a single-party minority cabinet of the Social Democrats, reliant on outside support from the Red–Green alliance.
Also, that second link notes the same point I made with respect to Israel: “Exit polls in Denmark can differ significantly from the final results.”
Fruits and Votes 
It almost sounds like Rasmussen has been watching a certain Danish TV series.
LikeLike
Indeed!
LikeLike
Apparently there’s some calls on the right for electoral reform after the Red Bloc won a majority of seats with a minority (49%) of the vote. Not sure on the details due to a paywall and a language barrier.
LikeLike
Interesting. I’d like to know more. Maybe I will ask on twitter, as I have a few Danish followers.
LikeLike
Here’s an unpaywalled version of the story. It’s a bit tricky to understand but I get the impression that Elkit thinks the system provided the Social Democrats with an “overhang” seat in North Zetland, giving them more seats in the districts than their nationwide proportional entitlement. Based on the Wiki map, the Social Democrats won no levelling seats at all. They won all of their 50 seats from the 135 allocated at the district level, giving them 37% of the seats for 26% of the vote. Evidently, the 40 levelling seats were (just) not adequate to compensate for that disproportionality. Elkit also notes that the left won two seats in Greenland, which has a lower enrolment per seat than the rest of Denmark.
LikeLiked by 1 person
My own calculations bear this out. Allocating the 175 seats amongst parties based on the Hare quota and largest remainders gives you 49 seats for the Social Democrats and 24 for Venstre. Had this been the final allocation, the total left bloc would have had 89 seats, one short of a majority. Had Denmark followed the New Zealand MMP approach of expanding the legislature to accommodate both the 24th Venstre seat and the 50th Social Democrat, the left bloc would have held precisely half of the seats (90 in a 180-member legislature),
LikeLiked by 1 person
Thanks. In the exchange I had on Twitter with a couple of Danish experts (Christensen and Paulsen), they mentioned that the debate is also over D’Hondt vs. Ste.-Laguë. The country switched from the latter to D’Hondt several elections ago. D’Hondt was consequential here in part because SD are large within their bloc while the right is fragmented.
LikeLike
Didn’t they switch to D’Hondt for the constituency seats while keeping modified Ste-Laguë for the leveling seats? I’ll have to find out but my Danish is rusty and there is unlikely to be much about it in English.
LikeLike
Yes this outlines it in English https://www.thedanishparliament.dk/-/media/pdf/publikationer/english/the-parliamentary-system-of-denmark_2011.ashx pages 6 to 8
LikeLike
But they moved to pure Ste Laguë from modified for the leveling seats — So from what I gather : Modified Ste- Laguë in both constituency and leveling seats to D’Hondt in the constituency and pure Ste-Laguë in the levelling. Happy to be proven wrong BTW. It’s a long time since I lived there.
LikeLike
No, Sainte-Lague isn’t used for the allocation of the levelling seats. The Danish system has four allocations: one at the district tier to allocate district seats between parties (D’Hondt), one at the nationwide tier to allocate total seats (and thus determine the number of compensatory seats) between parties (Hare and largest remainders), one again at the nationwide tier to allocate those compensatory seats to provinces (Sainte-Lague), and one at the provincial tier to allocate those provincial seats to individual constituencies (the weird Danish-only system with divisors of 1, 4, 7 et cetera).
I’m looking at the nationwide allocation of seats between parties, which is actually between multiple parties. It’s defined at Step Three of the system in the linked document. Sainte-Lague is used to allocate those compensatory seats to provinces, but this doesn’t actually impact the distribution.
LikeLiked by 2 people
So if they use LR-Hare for the overall allocation, I wonder how significant the divisor method used for the basic districts. One would think the compensatory seats would compensate, either way! I wonder if anyone has ever worked out a good simulation of how this matters.
LikeLike
The district-level system matters because in this case the number of compensatory seats is inadequate to fully compensate the Social Democrat haul in the districts. I’m not sure what would have happened had a Sainte-Lague system with a smaller district magnitude been in place (as was the case before 2006), given that such a system would be about as proportional.
LikeLike
And the modification was a 1.7 instead of 1 on the original count which gave a result somewhere inbetween D’Hondt and Ste-Laguë from memory
LikeLike
That’s interesting. Thanks.
LikeLike
I just checked the Bormann and Golder dataset of electoral systems. It shows a change in the district-level formula from “modified Ste.-Laguë to D’Hondt” between the elections of 2007 and 2011. However, it shows no change in the formula for the compensatory tier. Moreover, it shows that formula being “Hare quota and largest remainders.” So that is incorrect? I would think so, as that would actually be an odd approach to use for compensatory (leveling) seats.
LikeLike
Based on what Henry says above, apparently B&G have it right. Good!
LikeLike