General Synod Elections – Transfer Analysis Bath and Wells
This is an example of STV transfer analysis, undertaken on the Election for Laity members in Bath and Wells. Given that Bath and Wells published online their candidate’s election addresses, it has been possible to survey them and make assessments as to where each of the candidates broadly stood. I have, for the sake of simplicity, divided candidates into three camps. The first, signified by an E in the table below is a candidate who is a clear Evangelical or conservative Anglo-Catholic. The second, signified by an L is a candidate who is clearly standing on a platform similar to that proposed by Inclusive Church. The thrid category, signified by a C is a candidate standing on a centrist position.
I recognise that such divisions are rough but they serve the purpose of demonstrating transfer analysis. If anybody can provide more specific information on any of the candidates then please contact me.
The first table below shows the STV transfers.
You can see that on first preferences Armitstead and Hind were over the quota (63.67) and so their surpluses were reallocated. Then through the following rounds other candidates were either eliminated or passed the quota (having their surpluses reallocated). Wilson-Rudd was elected after the eighth stage upon reaching the quota, Humphreys and Stobart were elected by virtue of not being eliminated (there being 6 candidates left after stage 9). By my reckoning, one E candidate was elected, one C and three L.
The second table shows the Transfer Analysis and it is here that it becomes strikingly clear why the Ls gained three of the five seats. The top section of the chart shows you the percentage of voters who gave their first preference to a candidate of a particular persuasion. The second section shows you the way that voters who supported with their first preference a candidate of a particular “party” then used their second preferences.
31.9% of voters gave an E candidate their first preference, but then when their second (and third etc) preferences were allocated, only 27.3% of those E voters transferred to another E, whilst two thirds of them transferred to a C or an L. Compare this to the Ls – here almost two thirds of the preferences transferred to another L. This helps explain why the Ls gained three seats to the Es one – L laity voters in Bath and Wells were very careful not to help E candidates, whereas E voters happily supported L candidates. Maitland and Pearson simply did not get enough transfers from Armistead to survive.
I am unaware as to whether there was any organised advice on preference transfers given to L voters, but voting patterns like this seem to indicate as much. By comparison, E voters didn’t guard their E vote terribly well and ended up transferring support to C and L candidates. With more strategic transfer voting by E electors, Maitland or Pearson might very well have picked up the 5th spot instead of Stobart or Humphreys, given that the surplus of 42.33 from Armitstead added to either of their first preferences was easily enough to secure a second seat.
There may be a personal element involved here too: the top 4 candidates are pretty well known in the Diocese, and by Diocesan Synod reps, so there may be an element of â€˜go with what you knowâ€™ for voters.
Are the figures skewed by the fact that Edward Armitstead was the first one elected, so if there were any transfers to him from candidates 2 onwards, theyâ€™re not registered. Given that he was the most popular Evangelical candidate, any analysis of second preferences which doesnâ€™t include preferences for him is going to miss a sizeable part of the data.
Finally, I note that the net effect of the transfer system is that a L who would have got in on a straight top 5 is replaced by a C.
The answer to your second paragraph is “no”. The whole point of STV is that the votes are transferred to their “final destination”. If voters had put Armitstead as their 2nd pref, then once he’s past the surplus the vote transfers to the 3rd and 4th etc.
The real issue for Evangelicals is that many of those who voted for Armitstead as any of their preferences didn’t vote for the other two obviously Evangelical candidates. Now that might be becuase Armitstead had a large personal vote, but I don’t think that’s the full story, especially given the fact that the Inclusive candidates transferred votes to other Inclusive candidates highly effectively.
Peter, this is an enormously helpful and important piece of work. I tried to fathom out what was happening along similar lines with clergy voting in Chelmsford, but just couldn’t get my head around it. I’d be jolly impressed if you could do this!
Incidentally, I wonder how many ‘E’ clergy bothered to advise their Synod reps how to vote. I suspect most still feel this is either ungentlemanly or unnecessary!
If you send me a copy (pdf?) of the clergy election addresses I can have a go for you.
Oooh. I do like my picture! Episcopal purple?
You’ll be lucky…
If you go to http://www.gravatar.com you can change it to anything you like.