jmtd → log → vote
I've just voted Stefano Zacchiroli for Debian Project Leader (DPL). The voting period closes tonight, so if you are a Debian developer and haven't yet voted, please do!
Debian uses a Condorcet voting system, which to my non-expert eyes, bears some resemblance to "Single Transferrable Vote" (STV), at least compared to "First Past the Post" (FPTP). There is a debate raging in the UK as to whether Government elections should be counted based on an STV method instead of FPTP, with a referendum on the subject in May.
I'm leaning towards "Yes to AV", but this opinion piece by journalist Stuart Campbell (which argues that people should spoil their ballots) is worth a read.
Comments
My reasons for voting yes:
It (effectively) eliminates tactical voting, so people can vote for who they really want instead of having to think tactically about the effect that their vote will have. This adds to the legitimacy of the government, as the election is a more honest poll of citizens.
It requires MPs to have the support of at least 50% of voters, unlike in the current system where some MPs got in at the last election with only 25%. It will therefore strengthen the community link as they will have to work harder to keep their seats.
Votes are never “wasted” – the opinions of all voters are taken into account when deciding the winner.
Increases voting power index, which means increasing the influence of voters - see http://www.voterpower.org.uk/
Protects against extremist parties like the BNP coming to power, because they would have to get the support of a majority of constituents to win. The BNP have exploited the current system in some areas to gain council seats, because they only have to win the most votes – even though the majority of people voted against them. This is why the BNP are campaigning against AV.
Already used to elect the leaders of the Conservative party, Labour and the Lib Dems (Conservatives use multiple round runoff voting, which is conceptually the same).
Almost every other country in Europe already uses some form of preferential voting system. I'd rather have STV than AV but AV is certainly an improvement on our current system.
In-Response-To: 75YzuaX
AV (or as it's more properly known, IRV) is exactly the same thing as STV. It is simply the special case of STV where there is only one winner, so some of the more complicated STV parts like proportional vote redistribution after candidate reaches their quota are not needed.
As for IRV (AV) and any Condorcet method bearing any resemblance: the voting method sure looks similar (ranking any number of candidates in order of preference[0]), but the counting/choosing-the-winner methods are vastly different. IRV fails the Condorcet criterion, ie. it may choose a winner who would lose a one-on-one runoff vote with one of the other candidates.
All that said, IRV is a vast improvement over FPTP in reducing the scope[1] for tactical[2] voting, and that alone is a good enough reason for me to vote yes.
[0] Condorcet methods typically allow you to rank candidates equally, giving a partial rather than total order.
[1] But not eliminating entirely, I'm not aware of a voting method that can do this completely, though some are vastly superior to others.
[2] Or as it's more accurately known, insincere voting.