Predicting Regional Swing

First posted 22 August 2009

The simplest sort of prediction is called the Uniform National Swing which is a rather crude approximation and assumes no regional variation across the country. But using the results of a recent large YouGov poll, we can measure the extent of regional changes of opinion to get a more accruate prediction. This article describes how we do this and what the results are.

1. Regional opinion survey

The simple Uniform National Swing model assumes that the changes in support are uniform across the country. In other words, if the Conservatives gain 5% nationally, they will gain 5% in each region and consituency of the country. But the actual results are more complicated than this. In past elections we have seen non-uniform swing. For instance between 2001 and 2005, the Conservatives increased their vote share by 0.5% nationally. But within this average, there were different regional trends: they gained 3% in Essex but lost 1.7% in Yorkshire.

At the 2005 election, such regional swing accounted for 14 mis-predicted seats.

To make a more accurate prediction we need to estimate the size of these regional swing effects. But this cannot be done from national opinion polls alone. National opinion polls have a sample size of about 1,000 people, equivalent to 100 people per region, which is not enough to estimate opinion accurately. We need an opinion poll which surveys at least 1,000 people in each region of the country. Such polls are infrequent, but fortunately YouGov recently conducted one on behalf of Channel 4 for the European elections (YouGov poll details). The poll of 32,268 people was conducted between 29 May and 4 June 2009, and the headline results were Con 37%, Lab 22%, Lib 19% and Other 22%.

We are very grateful to YouGov for making the regional breakdown of their poll available to Electoral Calculus.

The table below shows (1) the breakdown of support by region as measured at the start of June 2009 by YouGov, (2) regional support as at the general election in May 2005, and (3) the changes from 2005 to 2009 which we call the Regional Swings. (Please see the notes at the end of this article to explain some small numerical discrepancies.)

(1) Regional Support June 2009
(source YouGov)
  (2) General Election Result
May 2005
  (3) Regional Swing
2005 - 2009
AreaCon %Lab %Lib %Nat %Oth % Con %Lab %Lib %Nat %Oth % Con %Lab %Lib %Nat %Oth %
Scotland21.027.013.030.09.015.839.522.617.74.45.2-12.5-9.612.34.6
The North32.031.020.017.019.552.923.34.312.5-21.9-3.312.7
North West32.028.017.023.028.745.021.35.03.3-17.0-4.318.0
Yorks/Humber33.025.019.023.029.143.620.76.63.9-18.6-1.716.4
Wales30.026.015.012.017.021.442.718.412.64.98.6-16.7-3.4-0.612.1
West Midlands40.023.018.019.034.838.918.67.75.2-15.9-0.611.3
East Midlands40.024.018.018.037.038.618.46.03.0-14.6-0.412.0
Anglia47.017.016.020.043.329.821.85.13.7-12.8-5.814.9
South West41.014.023.022.038.622.832.66.02.4-8.8-9.616.0
London40.025.018.017.031.938.921.97.38.1-13.9-3.99.7
South East50.013.020.017.045.024.425.45.25.0-11.4-5.411.8
Great Britain38.321.918.13.318.633.136.122.72.25.85.1-14.3-4.71.012.8

The national picture is that the Conservatives have gained about 5% support since May 2005, and Labour have lost 14%. The Lib Dems have lost 5% and Other parties have gained 14%.

Regionally, those trends are repeated in that the Conservatives have gained support in every region and Labour has lost support in every region. But although the direction of change is the same, the sizes of the changes are not. For instance, the Conservatives have gained only 2.4% support in the South West, but 12.5% in the North. Similarly Labour have lost only 8.8% in the South West, but 21.9% in the North.

2. Regional swing analysis

We can see these differences of swing in a more systematic way. Let us take the table of regional swings and subtract from it the average national swing. This will show us (4) the differences between the regional swings and the national swing.

(4) Regional Swing Differentials
2005 - 2009
  
AreaCon %Lab %Lib %Nat %Oth %Comment
Scotland0.11.8-4.912.3-9.2Strong SNP gain over LibDem and Others
The North7.4-7.61.4-1.1Strong Con gain over Lab
North West-1.8-2.70.44.2Others gain over Con and Lab
Yorks/Humber-1.2-4.33.02.6LibDem gain over Lab
Wales3.5-2.41.3-0.6-1.7Con gain over Lab
West Midlands0.1-1.64.1-2.5LibDem gain over Others
East Midlands-2.1-0.34.3-1.8LibDem gain over Con
Anglia-1.41.5-1.11.1Small Lab gain over Con
South West-2.75.5-4.92.2Strong Lab gain over LibDem
London3.00.40.8-4.1Con gain over Others
South East-0.12.9-0.7-2.0Lab gain over Others
Great Britain0.00.00.01.0-1.0

We remember that this is not a table of absolute swings. Absolute swings (table 3) show the Conservatives gaining everywhere and Labour losing everywhere. This is a table of swings relative to the national average. So, on average, everything balances out. For example, if the Conservatives perform more strongly than average in the North, they must do worse than average somewhere else, such as the North West. Thus the population-weighted average of each column is zero. (See notes below for why Nat and Others are different).

We see the following trends:
PartyRelative GainRelative Loss
CONNorth, Wales, LondonEast Midlands, South West
LABSouth West, South EastNorth, North West, Yorks/Humber, Wales
LIBYorks/Humber, West Midlands, East MidlandsScotland, South West

Overall there is a notable "depolarisation" as parties lose ground in their heartlands and make relative gains elsewhere.

3. Comparison with recent trends

We can compare these differentials with recent history. We have chosen the period 1992-2005 because Apr 1992 is the last election won by the Conservatives and May 2005 is the most recent election won by Labour. So this period captures the change in electoral geography due to the rise of the moderate New Labour and the waning of the Conservatives.

(5) Regional Swing
1992 - 2005
   (6) Regional Swing Differentials
1992 - 2005
AreaCon %Lab %Lib %Nat %Oth % Con %Lab %Lib %Nat %Oth %
Scotland-9.80.59.5-3.83.6-0.2-0.45.2-3.8-0.7
The North-10.5-1.07.83.7-0.9-2.03.5-0.6
North West-10.51.15.53.9-0.90.11.2-0.4
Yorks/Humber-8.6-0.53.45.71.0-1.4-0.91.4
Wales-7.2-6.86.03.74.32.4-7.81.73.70.0
West Midlands-9.3-0.23.95.60.2-1.2-0.41.3
East Midlands-10.31.34.44.6-0.80.30.10.3
Anglia-9.13.81.63.70.42.8-2.7-0.6
South West-9.13.21.64.40.52.2-2.70.1
London-13.41.96.05.6-3.80.91.71.3
South East-10.05.90.43.7-0.44.9-3.9-0.6
Great Britain-9.61.04.3-0.24.50.00.00.0-0.20.2

 

The regional trends over the period 1992-2005 are:
PartyRelative GainRelative Loss
CONWalesLondon
LABAnglia, South West, South EastNorth, Wales
LIBScotland, NorthAnglia, South West, South East

Interestingly, there is quite a similar pattern over the two periods 1992-2005 and 2005-2009. Labour continues to gain (relatively) in the south and lose (relatively) in the north of the country. The Conservatives continue to gain ground in Wales, though they have reversed their losses in London, perhaps due to Boris Johnson's election as mayor. The Liberal Democrats continue to gain ground in the north and lose relatively in the south. But in Scotland, their earlier gains have been reversed as the SNP increase their support.

4. Impact on Predictions

We can use these results to make a more detailed prediction of the next general election result. We can use both the national YouGov support figures, and the detailed regional support figures to make two separate predictions. We can then look at the difference between the two predictions to estimate the impact of regional trends on the result.

Party2005 SeatsUniform Prediction
June 2009
Regional Prediction
June 2009
Regional Impact
CON208371399+28
LAB346193160-33
LIB675150-1

The effect of the regional swing differentials in June 2009 is to give the Conservatives approximately another 30 seats at the expense of Labour. Running a similar calculation at August 2009 using the Regional Predictor showed the Conservatives gaining about 10 seats.

The Regional Predictor now has the added feature to make predictions incorporating the regional swings observed from this large YouGov poll.

Acknowledgements

We would like to express our gratitude to Peter Kellner and YouGov for making details of their regional poll available to us.

Appendix: Numerical Notes

YouGov themselves give some health warnings with regard to the regional support levels:

Additionally, Electoral Calculus gives its own cautions: