What would you call an electoral system that, under most conditions, proceeds in the following sequence:
(1) Take either the party or pre-electoral alliance that wins the most votes and give it an automatic bonus for being the largest;
(2) Then allocate the remaining seats proportionally.
With various country-specific features, that is how the Greek and Italian systems operate.
We lack a good name for it. Moreover, what seems to be the most common name given to such systems, both in the press and also by some political scientists, is “proportional”.
However, this is not an accurate name for a system that guarantees a disproportionally large share of the seats to the electoral entity that comes in first place.
So what should we call it?
It is really a hybrid of majoritarian and proportional, but of a fundamentally different sort than any mixed-member system. It aggregates votes nationwide (Greece, Italian Chamber of Deputies) or in regions (Italian Senate) and allocates a bonus or premium before any proportionality comes in to play. The name should reflect this premium on being the largest, as well as the openness to relatively small parties without regard to geographic concentration.
The orchard floor is open for naming suggestions.
Fruits and Votes 
What Alan said @49 (“… STV is proportional in terms of the final allocation of preferences…”) but with the further proviso that STV also tries to ensure – within the constraints of number of candidates over number of seats available – that final-count preferences match first preferences as closely as possible.
Eg, if a party with 4.9% of the votes fails to win a seat under STV, this is because there are fewer than 19 seats available to hand out and all the other parties have organised a tight exchange of preferences to lock that party out. Not because someone has decided that all parties under 5% get their votes thrown in the bin, even with a quota of 0.16667%.
LikeLike
By the way, for a textbook example of a “tight exchange of preferences” under IRO-AV see the House of Reps seat of McMillan in 1972 (http://tinyurl.com/a64w4nc and http://tinyurl.com/adrrxew), and under STV see the 1984 Senate contest in NSW South Wales http://tinyurl.com/ahrxao3, where Peter Garrett – lead singer of Midnight Oil, then candidate for the Nuclear Disarmament Party, now a federal Labor MHR and Minister – polled 9.6% on first preferences (with a quota of 12.5%, Australia’s so far only seven-vacancy half-Senate poll, to phase between ten and twelve Senators per State) yet lost after Labor and Coalition put him last. (Ironically, at the following Senate election in 1987 – admittedly a double dissolution – the NDP did win a seat in NSW with only 1.53% of first preferences, and without Garrett, because Labor relented and moved them up its ticket).
LikeLike
Another possibility is to all these systems “first place rewarded proportional representation”.
LikeLike
“PR subject to plurality bonus”?
LikeLike
After the performance of the system in recent elections perhaps ‘Potluck PR’ would be appropriate.
LikeLike
“No STV process I am aware of transfers preferences once all seats are filled and logically there is no reason to do so.”
Under Irish law, preferences are transferred if any candidate left still has a chance of recouping his deposit and expenses, which is a quarter of a quota. Most often, this means a candidate who cannot be elected and otherwise would be bloc excluded is not excluded to see if he will pass the deposit threshold. It’s very rare (I’ve not seen any examples) of this happening once all candidates have been elected, and it would probably require all candidates to be elected on the first count to happen.
I don’t think those preferences should be routinely counted. However, if you’re using it to determine funding (expenses or recouping a deposit) or else proportionality (be it deciding who is the ‘best loser’, as is currently the case in Malta, or if Malta were to use final preferences rather than first preferences), then I think it would be the right thing to do to distribute those preferences.
I think NSW has been pretty short sighted about not changing their surplus distribution rules thus far–they’ve had at least one referendum on the Senate voting system since inclusive Gregory was created (the one where they switched to single-party above-the-line tickets after 1999), and probably should have considered a change at that point. I don’t think a referendum solely on changing surplus distribution rules is a wise choice, but if another referendum is being called (say, to abolish below the line voting or something of that nature), I think it would be smart to include the change within it. It would probably be smarter to not entrench minute details like formality rules or surplus distribution in the constitution, and to simply entrench PR-STV with the details determined by statute.
“STV also tries to ensure – within the constraints of number of candidates over number of seats available – that final-count preferences match first preferences as closely as possible.”
With regards to Malta, this doesn’t always happen. Parties nominate wildly different numbers of candidates for the seats (the Nationalists nominated 4 candidates in one seat and 15 in another in 2008; both returned 5 MHRs). There is very little preference leakage to opposing parties (about 1%), but, especially when parties nominate large numbers, there are increasing numbers of voters who exhaust rather than ranking a further candidate of that party.
It seems to me that STV, or at least Malta’s variation, may not be well-designed for a very strong two-party system (Malta’s two-party dominnce is far beyond the “two-party” Labor/Coalition system of Australia, and has entrenched rather than reduced since the introduction of PR-STV). Malta uses 5-member districts, meaning one party will win more seats in every district, even if the split isn’t 60-40; they also use an odd number of districts, meaning one party will win more districts. The Maltese have been concerned with gerrymandering; very few seem to realize that it’s more the fact that the odd numbers and small magnitude give disproportional results. The combination of those two factors don’t always add up to the party with the most first preferences winning a majority of seats.
Buhagiar and Lauri 2009 talk about some of these issues in a pretty concise article from ‘Voting Matters’ on the future of STV in Malta here: http://www.votingmatters.org.uk/ISSUE26/I26P1.pdf
LikeLike
While it has been interesting, the discussion of STV procedures really should be somewhere in the block devoted to.. STV.
LikeLike
The recoupment of expenses rule in Irish elections should be regarded as something that happens after the election is complete. It does relate to electing TDs, only to deciding if a candidate should have their expenses recouped. In any case, as you say, it is a very rare procedure.
The odd number rule is generally regarded as a good thing. The ACT, which uses a much better entrenchment scheme than NSW, entrenches the principle that odd numbers of members should be returned in each electorate. And then there was the Tullymander.
This is a long way from the subject of the original post.
LikeLike
Grrr. ‘It does not relate…’
LikeLike
Pingback: The dismantling of SYRIZA? | Fruits and Votes