Sorry - should have been clearer. As STV would require all out elections and I'm only using the 2012 results, the comparison is with what the FPTP results would have been with multi-member all out FPTP elections in 2012, if that makes sense. The 2012 results were Lab 18, BIG 2, Con 1 which gives Lab 54, BIG 6, Con 3 (although arguably the Penistone wards and possibly Darfield were close enough they may have produced split results), so it's a fairly rough estimate but a better comparison than using three years worth of votes for FPTP when I've only used one for the STV estimate.
Pity that Scotland is not covered but the '2%' figure equates to Dummett's analysis. It may not be a high figure but assuming a whole council STV election one councillor in my borough (in Leeds two councillors) would owe their election to non-monotonicity.
Post by greatkingrat on Sept 13, 2013 20:29:19 GMT
I think it is 2% of wards, rather than 2% of councillors.
As the full preference breakdown has now been published for the 2012 Scottish elections, some clever person should be able to work out exactly which seats could have changed hands due to non-monotonicity.