16 April 2010

Tory / Labour seats / votes stats since 1979

Right. Have sobered up and checked the figures over. Hopefully this is at least vaguely visible. If not, the red line is the percentage of seats won by Tories / Labour in general elections since 1979, the green line is percentage of popular vote.

Two things to note:
1) The big bloody gap between them. Thank you FPTP. Particularly the 1983 drop-off in the popular vote having little impact on the seats gained.
2) Both lines drop over time, and the trend lines look on the better side of parallel.

I have estimated a 1.25 uplift from votes to seats. Now, if the share of the popular vote drops to 65% (reasonable given the more recent trend) then the share of seats should drop to about 80%, leaving about 125 seats in the hands of 'other. If - if - the vote share drops further, to 60%, the share of the seats would be c.75% and there would be about 160 'others' in Parliament (there are 92 at the moment, the most ever).

Anyway. Stats not really my thing and I'm crap at spreadsheets, but this is what I was wittering about last night after the debate.


  1. Great stuff, roll on "Other"
    (NB technophobes, click on the graph and it will enlarge for better visibility).

  2. Roll on other - yes depends upon who other is doesn't it ?

    How much of the increasing discrepancies are accounted for by the increasing numbers of independent candidates and changes in electoral boundaries ?

  3. Leni - the figures changed a bit when I sobered up and managed to find the total seat numbers...those mainly change for redrawn boundary lines. I have probably missed something big relating to N Ireland, but don't have the strength to go digging back around, as I'm not a good enough statto to make the spreadsheet any easier!

    Wanted to put in another trend line from 1992, as that would be steeper, and extend to a 65% vote share (I think) but can't work out how to miss out the earlier years...if I'm the UT numbers whizz, we're in trouble...

  4. Philippa

    I'm no expert but wouldn't you have to do separate graph as you would be starting from a different place ? Or assume the change in vote share reflected voting habits of transferred part of population .

    That might be rubbish -- however.

    I would like to see what happens if Cameron gets in and reduces seat numbers by 10%. Luckily from a post election stats approach you have time to work on that one . Would be nice to see proposed new boundary lines and work out implications before this coming election. No chance of that of course.

  5. Hi Philippa

    Would be really interesting to see just how
    different the political landscape would have
    been if we had had PR since 1979.And obviously
    if/when PR is introduced in the UK the 'other'
    category which may win seats will include the
    BNP -as in the Euro-elections-so there is a
    downside!Cheers for posting this!

  6. Ooh, Paul, there's a challenge.

    Will return to spreadsheet, and make someting up, sorry, try modelling a different approach...