Lesotho and Malta will hold early elections this Saturday, June 3rd. Both have parliamentary systems and each one uses a different (and interesting!) type of proportional representation – each having a certain following among readers of this blog.
Lesotho uses a one-vote variant of MMP, with 80 single-seat districts in the nominal tier and 40 in the list tier. There is no threshold, and no seats are added in case of overhang, so a party can win a majority by taking more than 60 districts.
Malta uses STV, with a twist: if I understand correctly, in case one party receives an absolute majority of first-preference votes, seats are added to ensure that party has a majority, and that the majority is in proportion to its majority of the vote.
The elections were also called in different ways. Lesotho’s parliament (election not required before February 2020) was dissolved after the government lost a confidence vote in March – the prime minister could have handed over power to the coalition that ousted him, but chose instead to ask the king for an early election. Malta’s early election (originally not due until March 2018) was called by the prime minister.
Fruits and Votes 
If there are no provisions for desaling with overhang seats, which party or parties end up short if party wins “too man” districts?
LikeLike
Basically, all the others.
The calculation goes as follows: tabulate row of proportional seat shares (PR), row of district seats (DS), and take away the latter from the former. e.g.
Party A B C
PR 40 40 40
DS 60 20 0
—————–
-20 20 40
The last row should be the number of list seats. However, you can’t give ‘negative’ list seats. So just exclude party A from the calculation. Now you’re left with:
Party B C
PR 30 30
DS 20 0
—————–
10 0
So the final tally is:
party, total seats (district, list)
A 60 (60, 0)
B 30 (20, 10)
C 30 (0, 30)
LikeLike
Small typo… middle table should be:
Party B C
PR 30 30
DS 20 0
—————–
10 30
LikeLike
I have also read that Italian parties had agreed to a deal to change their electoral system to that of “German style MMP” with a 5% threshold. Is a new post on that with local sources coming up?
LikeLike
I’ve seen, in English sources, references to a “German style system” with a 5% threshold. My first instinct is MMP, but my second is that it could just as easily be closed list proportional, or a return to their psuedo-mixed member system of the past. I’m also getting a real sense that whatever the actual system is, it is just an attempt to prvent a 5 Star government by eliminating any plurality bonuses.
LikeLike
Maltese STV has a second twist: if only two parties win seats in Parliament, top-up seats will be distributed so that the seats are, as closely as possible, proportional to the number of first-preference votes received by each party (I can’t recall the exact rule used to determine exactly how many seats as I haven’t read their election code in some time).
In the last election, in 13 5-member seats, Labour won 39 seats and the Nationalists won 26 seats. The Nationalists received 4 top-up seats to make the result proportional, for a final split of 39-30.
One of those districts, the 13th (which includes the islands of Gozo and Comino and is much less populated than the other 12 districts), had a plurality reversal. The PN narrowly won more first preference votes, but due to ballot exhaustion, the MLP won 3 of the 5 seats.
I am not sure whether the top-up rule will be affected by the fact that the main opposition to the Labour Party in this election is the new Forza Nazzjonali, which is technically a coalition between the Nationalist Party and the Democratic Party (a centrist Labour breakaway). I don’t know whether the fact that it is a coalition and not a party would impact whether the winner or loser receives top-up seats; the AD Green Party, though clearly a minor party, has come very close to winning a seat in one of the districts in the past two or three elections, and this would also impact the final distribution of seats.
LikeLike
Exhausted votes cannot cause a ballot reversal. STV results are proportional to the final distribution of preferences, not to the first preferences. South Australia has a ballot reversal rule for its assembly, although it is somewhat ineffectual because it is an instruction to the boundaries commission, rather than a topup arrangement after an election. Their language is:
I have argued for some time that every districted electoral system needs a fallback clause to deal with reversed majorities. The South Australian language would be much better than the Maltese reliance on first preferences.
LikeLike
The problem with the South Australian gerrymander is that it doesn’t prevent a reversed plurality from governing, it merely lessons the chances of that government securing a majority at the next election.The current and inefficient distribution of liklely Liberal voterss give them strong majorites in a minority of divisions, and at the last election the two-party prefered majority, while Labor wins enough votes in the marginal seats to keep power.
As face saving methods go, “Give the Party that Got a Majority Enough Seats to Have a Majority” is probably the best method. It just requires, in Malta’s case, that first preferences be considered the most important part of the vote. The system would, I imagine, break down if cross-party preferences became a thing and one party’s votes giving a seat to the other party’s candidate caused the reversed plurality.
LikeLike
I fully agree that the Gozo situation is not a true plurality reversal, but the Naltese election law essentially treats it as such.
It appears that the Forza Nazzjonali has been treated as a single party and awarded top-up seats. Labour still won the election convincingly and lost just a single seat despite significant corruption allegations.
It’s also worth noting that in Malta, candidates are permitted to stand in two districts, and top candidates frequently win two seats, in which case they choose one and the other seat is awarded on countback, though I can’t remember the specific countback mechanism off the top of my head.
LikeLike
Also, from what I understand cross-party voting is quite rare in Malta (due to societal pillarization among other factors), and a reversal of a first-preference plurality is much more likely to come from exhaustion.
I do think it’s well-known in Malta that the first preference vote is considered to be voting for that party for seat distribution, though I don’t have any evidence to back up that belief.
LikeLike
Mark, it is simply wrong to describe SA as a gerrymander. The boundaries commission is independent. The redistribution standards are fair and reasonable. The redistribution process is open and fair. The courts can review a redistribution on application by any elector. The margin of varPoliticians have no role beyond making public submissions to the commission. If single member districts were going to work anywhere it would be South Australia, but they just don’t work.
LikeLike
I would certainly describe South Australia as a gerrymander. It is simply not a deliberate, negative gerrymander. It may even be the most positive gerrymander ever. But it is still an attempt to “manipulate the boundaries of (an electoral constituency) so as to favor one party or class,” in this case the party being the party recieving a two-party preferred majority and the class being the voters that created the two-party preferred majority.
I understand why South Australia does what it does. I too want to balance regional interests and the at-large preferences. (A far better system would be to simply add top-up seats ala Malta if needed) But to deliberately skew the idea of a local member by drawing boundaries to favor a preselected favored group is Gerrymandering.
LikeLike
Gerrymandering is usually taken to mean a deliberate process. There is simply no evidence of gerrymandering in SA. In particular there is no evidence of drawing electorates that favour sitting members, nor would that would be permissible under PART 5–Electoral redistribution of the state constitution.
LikeLike
We’re starting to get some results!
Malta remains as two-party as ever. In fact, both major parties’ vote shares are a tiny bit up on last time. Labour is on 55%, while the Nationalists are on 43.7%. No other party broke 1%.
Lesotho’s count is about 70% done. It won’t be clear how close the biggest party will be to a majority until all the constituencies have been counted, however. The problem is that in three seats, the candidate died before the election, and so by-elections will be held in a few months – by-elections whose tallies will have to be added to the rest to calculate the total number of seats per party! So we won’t have final results for at least another three months or so.
LikeLike
Technically Malta has a 3rd party which won seats (the first time a third party has won seats at an election since 1962 iirc)–the Democratic Party won at least one seat as part of the Forza Nazzjonali alliance. They appeared on the ballot as “Nationalist Party–The Orange Ones,” while most Nationalist Party members were just listed as “Nationalist Party.”
However, the PD’s leader is now talking about running for leader of the Nationalists, so this may not be the case for long.
LikeLike
Interesting. So the two members of the alliance actually had separate columns? How many candidates did they each run?
LikeLike
No, they were in the same column but the PD candidates were marked, as far as I can tell.
The nominations wildly differ depending on the constituency. In Malta, the major parties have more candidates than there are vacancies (I am not sure how candidate selection works or whether the parties have any control over who runs over their banner–I’ve been looking for a source for some time).
For instance, in this year’s election, in the 1st district, which like all others has 5 seats, the Labour Party had 9 candidates, Forza Nazzjonali had 9 (8 PN and 1 PD), and others had 5. Within each party/alliance, candidates are, I believe, listed in alphabetical order, so the PD candidate was 2nd of 9 on the Forza list.
Labour had a low of 6 candidates and a high of 14. The Forza Nazzjonali had 21 nominees in one district, of whom 19 were PN and 2 were PD.
LikeLike
There is an image of the 2014 European Parliament ballot here (as well as for the rest of Europe), but it was before the Labour Party/Democratic Party split iirc.
https://www.kai-friederike.de/EP2014_ballots.html
LikeLike
That is an excellent resource, Chris, thank you.
LikeLike