Saturday, 2 March 2013

Eastleigh & AV

This week has seen the Liberal Democrats retain the seat of Eastleigh.

One response is to note that a centre-right vote has been split allowing the Liberal Demcorats to hold on. That is probably a bit simplistic, and as the study by Conservative peer Michael Ashcroft shows, there were a lot of cross-party dynamics.

One cannot simply assume that people who voted for party X would have voted for party Y if Y had had one of X's policies.

And you should never assume that if only Y adopts X's policies then it can do so without losing any voters. In the Daily Mail, we see Simon Heffer fall into the trap of stating that:

It has been estimated that [the UK Independence Party] prevented the Tories from winning up to 20 more seats in the last general election. How many more will it deny them in 2015?

These estimations rely on various assumptions:

  • That people who voted UKIP at the May 2010 general election were drawn nearly exclusively from disaffected Conservative voters
  • That adopting policies pleasing to UKIP would bring these voters back to the Conservatives without alienating any existing Conservative voters
  • Hopefully, one outcome of the Eastleigh by-election will be the recognition that UKIP is an anti-politics party, taking the traditional protest vote role from the Liberal Democrats, and drawing support from all parties. We should see the end of the simple "add the Conservative and UKIP vote to get how many votes the Conservatives would have got if they had traditional policies".

    What if the by-election had been under the Alternative Vote system? The real vote was:

  • Mike Thornton (Liberal Democrats) - 13,342
  • Diane James (UKIP) - 11,571
  • Maria Hutchings (Conservatives) - 10,559
  • John O'Farrell (Labour) - 4,088
  • Others - 2,056
  • The first stage would have been to delete all but the top 3 candidates. This might seem drastic but if all the voters for the Others gave their second preferences to O'Farrell, he would have 6,144 votes, and would be 4,415 votes behind Hutchings. Hence, he would not win under AV and can be deleted.

    Now, how would the vote go? If we look at the summary of the Ashcroft poll, then we see that out of those who voted Labour in 2010 and who voted on Thursday, 23% voted for Thornton, 17% for James and 6% for Hutchings.

    There are a couple of things to note:

  • Labour didn't stand a chance of winning in Eastleigh. Although at the July 1945 general election (when most of the current Eastleigh was in the Winchester constituency) Winchester had a Labour MP (George Jeger) and at Eastleigh's inaugural general election of May 1955 it was a close Conservative/Labour marginal (and Labour came close to winning in March 1966), the June 1983 general election was the one where the Liberals dramtically overtook Labour to become the main non-Conservative party.
  • In the light of this, there is significant tactical voting for the Liberal Democrats despite being in coalition with the Conservatives.
  • It is reasonable to assume that around half of Labour's second preferences would go to Thornton, and the remaining half split between James and Hutchings, but with James getting more than Hutchings so extending her lead.

    Although supporters of the Others might rank Hutchings above Thornton and James, this would be unlikely to be enough to avoid her getting knocked out at the second stage of counting, leaving the third stage of counting being the battle between the Liberal Democtrats and UKIP.

    From the Ashcroft poll, out of Conservative voters in 2010, 22% voted for James and 14% for Thornton.

    Put some figures on this. I will:

  • Ignore the votes for Others
  • Assume that second preferences among remaining candidates would be distributed in proportion to the change in allegiance in the Ashcroft poll
  • In practice, let's start with Labour's 4,088 votes. As remarked about, out of those who voted Labour in 2010, 23% voted Liberal Democrat, 17% UKIP and 6% Conservative. That makes 46% in all:

  • Liberal Democrat - 23% / 46 % x 4,088 + 13,342 = 2,044 + 13,342 = 15,386
  • UKIP - 17% / 46% x 4,088 + 11,571 = 1,511 + 11,571 = 13,082
  • Conservative - 6% / 46% x 4,088 + 10,559 = 533 + 10,559 = 11,092
  • Finally, the Conservatives' 11,092 votes. As noted above, out of those who voted Conservative in 2010, 22% voted UKIP and 14% Liberal Democrat. That makes 36% in all:

  • Liberal Democrat - 14% / 36% x 11,092 + 15,386 = 4,314 + 15,386 = 19,700
  • UKIP - 22% / 36% x 11,092 + 13,082 = 6,778 + 13,082 = 19,860
  • To be honest, too close to call - depends on whether supporters of Labour and the Others would prefer the Liberal Democrats or UKIP - but it is possible that under AV, James would have become UKIP's first MP.

    No comments:

    Post a Comment