And here's the patronising from exponents of the broken FPTP. "PR is really really complicated, so it's easier to stick with the much simpler FPTP that everyone understands".
Apart from all the different possibilities for tactical voting, some of which work and some of which are only believed to work. That's why I love electing my Federal and State MPs using the Instant Runoff Vote (aka AV in the UK) where I can vote for who I want first, who I'd be happy to have in second, who I'd put up with in third and who'd be the lesser of two evils in fourth.
You would be hoping that Labour would be changing their tune on real electoral reform after this, they were the big losers of the system this time around.
And here's the patronising from exponents of the broken FPTP. "PR is really really complicated, so it's easier to stick with the much simpler FPTP that everyone understands".
PR is not complicated. The methods used in the counting process are probably more complicated, but this is needed to ensure proportionality. The results give us proportionate and democratic representation to parliament, which clearly does not happen under the current system. To the voter, little will change. You'll still mark your ballot with an X, or rank your preferences with numbers. Neither is beyond the realms of comprehension.
Indeed. There are different forms of PR with varying degrees of complication.
The 'simplest' one which would also continue the legacy of local members in the UK would be to use an Instant Runoff Vote (also known in the UK as AV) for roughly 450 constituencies with an approximately equal number of electors (in Australia this is legislated as +/- 10% and regulated by the non-partisan Australian Electoral Commission) in each, and using the same votes to elect additional members. The percentages of the national (i.e. UK-wide)
primary vote won by each party could then be used to add 100 extra non-local members to those parties which wind up with fewer seats than they would "deserve" as a result of their share of the national primary vote.
The strength of using the 'local' primary vote for the additional members is that you place local representation first, and party representation second. A party which does well enough to win more seats than their share of the national vote entitled them to (e.g. the SNP, or an Independent) would still get to keep all the seats they earned, they just wouldn't get any additional members.
How to pick the additional members is the sticking point with such a system - I'd be in favour of the system used in some states of Germany where it is the highest-polling 'losing' candidates of the parties, which ensures that a decent level of support (but not quite a winning level) is needed to get in as an additional member instead of just being a party hack on a list. This could have the effect of minimising the impact of a select number of marginal seats on the campaign strategies of the parties - if it's going to be a 50/50 seat then your candidate will get in either way.
You can stand as an independent under PR, and it's been done before by Margo MacDonald here in Scotland! You'd have have appeal to a larger voter base in order to get elected, but as the swings towards the two big parties make it more difficult for independent candidates to get support under FPTP, it's probably easier to get in under a proportional system, not least because the electorate can support independent candidates without fear that their vote will be wasted.
Hare-Clark (the form of STV you prefer) is used to proportionally elect six Senators from each State of Australia at each half-Senate election, and allows for both independents and minor parties to gain seats. South Australia currently has an Independent Senator in the form of Nick Xenophon, who made history by being the first Independent candidate for Senator from any state to gain in excess of a full quota on primary votes alone at the last election, and at the previous election became the first Independent Senator ever to be elected without having first jumped ship from a party.
An additional member system could also honour Independents properly, if it allowed votes cast for Independent candidates to have the voter's
second preference counted towards the national total for the purpose of electing additional members - maybe at half value only if the Independent got elected. Under this, a person could cast their vote for a strong local member first while also expressing their hope for the national-level policies of a party.
A good example of the sort of seat where this would be useful would be the Edinburgh South constituency where the incumbent Labour MP Ian Murray was comfortably returned on the back of being an excellent local member, with even the SNP party volunteer interviewed by the ABC admitting to voting for him. Fair play to him, even SNP supporters should be glad at such a victory for democracy and hope that their newly elected MPs will look to him as an example of how they should serve their constituents. Change 'Labour' to 'Independent' and you can see the value of such a system.
I'm a proponent of the Single Transferrable Vote, and here's a video which demonstrates it very nicely.
[youtube]l8XOZJkozfI[/youtube]
It's difficult to say how PR would have worked in this election, but to give you an idea we can look at some seats. Let's look at the seven Glasgow constituencies (that is, the constituencies with the word "Glasgow" in their name), each of which returned an SNP MP, and merge them into one large constituency to return seven MPs. I can't simulate an accurate election here because we don't know what second and third preferences are, but here's how things could have looked in the first round of a Greater Glasgow PR style Vote:
Electorate = 252,210
Quotient (number of votes needed for election) = (252210/7+1)+1 = approx 31,527
SNP: 163,937
Lab: 91,280
Con: 20,323
(And a plethora of smaller parties that I can't be bothered adding up)
In this model, SNP and Labour are automatically elected in the first round. We'd then go through a large number of run-offs to transfer/redistribute votes to 2nd and 3rd preferences, which I can't simulate here without knowing data that we'll never be able to know, but it'd be likely that we'd see something like 4-5 SNP and 2-3 Labour MPs, which is closer to what the people of Glasgow as a whole voted for. The vast majority of the electorate are represented by at least one local candidate from a party that they voted for.
This model assumes that:
1) Everyone who voted would vote again under the new system, and no additional votes would be gathered
2) Everyone voted for their first choice candidate in yesterday's election, and would rank that candidate as their first or only choice in this election.
I'm a fan of Hare-Clark (the form of STV you support) for an upper house (e.g. the Senate in Australia) where you have large regions electing multiple members (e.g. a whole State) but not for a lower house as local members are still worth having.
For a Senate-style elected House of Lords (keeping the name for tradition's sake) elected using Hare-Clark, the UK would probably be best off by having multi-member constituencies using the same boundaries as the European Parliament constituencies with NI, Scotland and Wales as one each and England divided into nine. If the number of Senators/Lords from each was double the current number of MEPs plus two, you'd also get a bias towards regional equality (the extra two would be an extra 10% for SE England, an extra 17% for Scotland and an extra 33% for Northern Ireland) while still being roughly equal by population.
Going on the raw numbers of first preferences, you would probably be looking at SNP 5 and Labour 2. But re-running the election without tactical voting (which made up a small part of the SNP's gains) would probably change it to SNP 4-5, Labour 1-2 and Con getting the last seat on preferences. From memory, the last half-Senate election in Australia (where we are used to this system by now) had none of the six states electing any more than four of the seats on quotas being achieved just from first preferences, it does have the effect of opening up the field a bit.
To work properly, in my opinion it needs to require that the candidates' positions on the ballot paper be allocated at random and not grouped by parties - the Senate ballot papers here have an option to use the party's preferred allocation of preferences which is not optimal as the system can be gamed. The use of the Robson Rotation to randomise the advantage of being positioned first on the ballot paper would be desirable too.