Monday, April 26, 2010

Election Outcome Scenarios for the UK Election

I use past elections as a guide and supplement it with polling results.  Polls in the UK suffer from the same problems they do here in Canada, many more people express an opinion than actually will vote.   UK polls are getting 85-90% of respondents saying they will be voting which is highly unlikely.   There is a large built in error in all of the results because of this high number of respondents that are giving an opinion but will not be voting.   In most elections the error does not seem to be a factor, but in an election like this one where there has been a major change in what voters are thinking of doing, it will matter dramatically.  I will factor this into the scenarios

All three scenarios are within the range of what the polls have been saying for the last 12 days.  I am also using the best regional numbers I can get, but I have to say they are weak in the UK.

Scenario One Assumptions:
Polls are broadly accurate
Increase in voter turnout 67% with a total vote of 30 million
Conservatives  33.3% - 10,000,000 votes
Liberal Democrats 30% - 9,000,000 votes
Labour  26.7% - 8,000,000 votes
All other parties 10% - 3,000,000 votes
In this scenario the Conservative vote rises, but no faster than their share of the vote.

  • Party            Number of Seats
  • Conservatives - 270-290
  • Labour 140 - 200
  • Liberal Democrats - 130-150
  • SNP/PC - 12-15
  • Ind + Speaker - 2
  • Northern Ireland - 18

Labour has the most volatility in their vote because it is clear that it is them and not the Conservatives that are losing votes.   The swing will not be uniform, one of the impacts of the 2005 election was that Labour cratered in vote in many seats across the country, there are not many votes left to lose in places like Bath or Woking.  Labour will end up losing more votes in seats that they won in 2005 than anywhere else.  Their margins of victory are smaller than the other parties and their seats are in much danger of being lost.

In this scenario there is a chance that the Liberal Democrats will gain more seats than Labour, but it is not likely.

This scenario leaves the Conservatives far enough from a majority that they need the support of one of the other major parties.

Scenario Two Assumptions:
Polls over estimate Liberal Democrat and Labour support
Voter turnout rises marginally to 28,500,000
Conservatives 35.1% - 10,000,000 votes
Liberal Democrats 29.8% - 8,500,000
Labour 25.8% - 7,350,000 votes
All other parties 10% - 2,850,000 votes

  • Party     Number of Seats
  • Conservatives - 320 - 360
  • Liberal Democrats - 120 - 160 
  • Labour - 110 - 170
  • SNP/PC - 14-16
  • Ind +Speaker - 2
  • Northern Ireland - 18

Labour still suffers from the most volatility because of their position as the main party dropping in support, but with a weaker vote for the Liberal Democrats, the Conservatives are in majority territory.   In this scenario it is likely that Labour and the Liberal Democrats will be close to even in seats

Scenario Three Assumptions
Polls over estimate Labour support
Vote turnout rises to 29,000,000
Conservatives 34.5% - 10,000,000 votes
Liberal Democrats 31.0% - 9,000,000 votes
Labour 24.1% - 7,000,000 votes
All Other Parties 10.3%  - 3,000,000 votes

  • Party         Number of Seats
  • Conservatives 280 - 320
  • Liberal Democrats  200 - 250
  • Labour 50 - 110
  • SNP/PC 15 - 20
  • Ind + Speaker 2
  • Northern Ireland 18

This is what I think will be the most likely result on election day.   I expect Labour to suffer from a poor turnout on election and this will translate into a lot of seats being lost and with the majority of them being won by the Liberal Democrats.  

This result means that it is Gordon Brown and Labour that are the potential kingmakers in the next parliament.  Nick Clegg is the clear leader of the Official Opposition and is the PM in waiting.  

If David Cameron is at more than 310 seats, he should be able to govern with some confidence without a formal partner.   If he ends up with fewer than 300 seats the parliament will be much more like the minority parliaments in Canada, Cameron will have to govern with the tacit support of the main opposition in parliament.

Scenario Three is Most Likely in My Opinion
Why do I think this outcome is most realistic?   It comes from watching numerous elections in Canada with dramatic voter shifts in elections.   The big impact on election day for the parties that suffered badly came from their supporters staying home on election day.   The impact of this is that all the close races are obviously lost, but a surprisingly high number of 'safe' seats are lost as well.

As an example, if you are normally a Labour supporter in a place like Lewisham Deptford, you will have been lackadaisical about voting unless you are really partisan.  Most people in this seat have not been voting, but Labour has still won with large majorities.   Labour supporters could very well feel that them not voting will have little impact in the election.   It is places like this that provide the real surprises.

In my opinion the pundits in the current UK election are too conservative in their thinking when it comes to how many seats Clegg can win and more importantly, how many Labour can lose.


Anthony said...

Hi Bernard,

Can you say anything more about the specific mechanics you used to generate these scenario outcomes?

Bernard said...

I will try to, I have a spreadsheet on my comp.

1) Most voter turnout increase was applied to seats with less than and 50% voter turnout in 2005. Most of this increase was assigned to LD

2) Labour vote was reduced primarily through the following split - 30% to LD, 15% to Cons and 55% to non voting. High voter turn out seats in 2005 were only assigned a small increase in turnout.

1) and 2) the attempt was to emulate the sort of results we saw in Canada for Conservatives in 1993 and Liberals in 1984. Some emulation was done of UK Conservative results in 1997

3) 2005 Labour marginals in which Cons were close were all assigned to Cons.

4) What regional breakdowns I could find were applied.

5) Incumbency was assigned as a bonus to all 1 term LD and Cons but only to well know Labour MPs

If I get some more time, I will explain in more detail in a posting

Anthony said...

Interesting, thanks,

Unknown said...

I live in Lewisham Deptford, and I'm going GREEN this time....