Bristol’s Mayoral Election: A statistical perspective
Elections allow those politicos who like to play with figures a fantastic opportunity to prove that the losers actually won and that their team did fantastically. I can’t resist an analysis of my own but will try and remain as objective as possible. I have looked at two issues:
1) How would the parties have expected to have done if there hadn’t been the huge vote beyond the mainstream parties (including the Greens).
2) What are the long term trends in Bristol based upon the last 4 all out elections in the City.
How well did the parties in Bristol really do?
The last election which all people in the city could vote? That was the general election in 2010. The turnout then was just over twice that on Thursday. 196,077 votes compared to 89,156. This means that any comparison is affected by two factors. That the profile of the turnout across the city is not the same and that in a general election people may vote differently than in a local election. A third factor is that the voting system is different allowing people to have a heart and a head vote in Thursday’s election whereas in a general election some people will vote tactically for the candidates who they think might be in with a chance winning. This means that in Thursday’s supplementary vote election you would expect ‘minor’ parties to do better.
The Results for the two elections were as follows:
# 2010 2012 Change
Liberal Democrat 34.3% 7.0% -27.3%
Labour 32.3% 29.0% -3.3%
Conservative 27.1% 9.1% -18.0%
Green 2.4% 5.9% +3.5%
Left parties 1.1% 3.3% +2.2%
Others 4.3% 45.6% +41.3%
However the national polls are not the same as they were at the general election so to have a more accurate picture of party performance we need to adjust for the change in the national picture. I have done this taking the poll of polls published by Anthony Wells on his yougov related blog.
# 2010 %s 2012%s Change
Labour 29% 42% +13%
Conservative 36% 33% -3%
LD 23% 9% -14%
Others 12% 16% +4%
If we factor these changes into Bristol (which assumes that the national changes are reflected in Bristol) the percentages we should have expected in the mayoral election and the difference to the actual result are as follows:
# Projected Variation
Labour 45% -16%
Con 24% -15%
LD 20% -13%
Others 11% +44%
This tends to suggest that the party which did worse compared to what might be expected was Labour (although rounding means it could be just as bad for the Conservatives). A reading of the turnout statistics suggests that much of this loss was due to spectacularly low turnouts in traditional labour areas (barely above 10% in Labour’s core areas while nearer 40% in conservative and lib dem ‘core areas’ – excluding postal votes) but it would be hard to argue that the whole effect was just due to this and that some normally Labour supporters did vote for independent or ‘minor party’ candidates.
Long term trends
Bristol rarely has all out elections. This is due to the ‘election by thirds’ in local elections. Also prior to 2010 the parliamentary boundaries did not match the city boundary. There were two all out elections in 1999 (following boundary changes, now long overview) and 1995 following the abolition of the County Council. 1995 was a the high water mark of the Labour vote, at least since the formation of the SDP and the last time that one party got over 50% of the vote in the City. This does skew the analysis a little but the trends in party share in the 4 all out elections does tell an interesting story.
Labour appears to be on a steady slide, The Lib Dems and Conservatives have seen a catastrophic collapse in their vote last week and the greens have risen (partly due to the difference in voting system) but are still trailing the other mainstream parties.
It could be argued that this recent vote was unusual, which it certainly was. It we look just at the four parties share ignoring other parties the graph looks very different:
This shows Labour in a strengthening position after Thursday in comparison to the Liberal Democrats and Conservatives.
It doesn’t take a genius to realise that the election result was bad news for the three traditional political parties. The analysis here suggests it was much worse for Labour than just looking at the votes cast might suggest. However it does also suggest that if we start to see more independents standing in local elections following this result Labour may be in a stronger position than the Conservatives and Liberal Democrats in holding and gaining council seats.
Labour will be hoping that turnout rises among its core voters in future elections and the Tories and Lib Dems that after a brief flirtation with independents that their voters will come back to them in May. The Greens will be hoping that some of the independent vote will come to them and that they can hold onto the voters who want to vote for them but don’t wish to see their vote wasted.
Party politics in Bristol’s local government is not dead, indeed traditional political parties (including ‘minor’ parties) polled more than independents (just) but it is wounded.