Forecast #GE2017 – 7th June 2017


This is our latest forecast for the 8th of June 2017 General Election in the UK.

Will your forecast change between now and tomorrow? – Very likely. We expect a burst of polling activity over the next 24 hours and we will produce one or more forecasts tomorrow to reflect updated information.

What does your forecast show? – We try to give as much information as possible, but only where it is statistically significant. We try to show a 50% confidence interval on the national and regional vote share for the major parties and the range of likely seat possibilites. In doing the latter we show a much narrower range of possible seats for each party than some other forecasts as we believe a wide range of seats, for example the wide range used in the YouGov model, makes the concept of a “prediction” almost meaningless. The range of seats we indicate is the narrowest one which we believe is possible to offer and is based on the mid-point of our forecast. It is the “most likely” outcome but we accept that it does not cover all possibilities.

What do you use for your forecast? – We use all the latest polling data, where available. We also look at the betting markets and other information to help guide our forecast. We calculate the interaction between the support for the parties on as local a level as statistically possible and then use this to run a Monte Carlo simulation of the election.

What do you predict will happen in Seat X? What is the probability of Party Y having more votes than Z? – As we approach the 8th of June we reduce uncertainity in our model and are able to answer questions like this. Whilst we do not automatically publish a prediction for each seat, we can indicate a most likely outcome if required and also probabilities of victory for each candidate.

I want to ask a question / get in touch – Write a comment below to get in touch.

Key Features

The overall summary of our forecast is “Modest Gains for the Conservatives”. Most of the current volatility in our forecast is caused by uncertainty in our Scottish and Welsh forecasts.

My now include a forecast for London. Labour is expected to perform much better in the capital than elsewhere in England.

Our forecast began to move towards the Conservatives after the weekend’s polling. We continue to examine the issues around turnout weighting and note with interest that the two companies with the most sophisticated turnout models, ComRes and ICM, have not only produced the largest Conservative leads but are also clustered together in their results.

Our Scottish and Welsh forecasts now have expected vote ranges in each of the regions. We continue to see evidence that all three national parties will make gains from the SNP on Thursday in Scotland.

Our Welsh forecast (as indicated by the large vote ranges displayed) has a high degree of uncertainty in it and we are cautious as to the actual result. This caution and uncertainty is driven by the one recent large survey (YouGov) being contradicted by national poll subsamples. Where (as in Scotland) we rely on a number of large samples alongside sub-samples, the range of uncertainty is much narrower).

We explore the probability of the Conservatives and Labour beating each other for vote share in Wales (for first place) and Scotland (for second place). Both probabilities show an absolute neck and neck race.

Our forecast indicates that there is little sign of a Remain boost for the Liberal Democrats. We are predicting a slump for UKIP and our seat forecast shows them not picking up any seats.

Our Northern Ireland forecast is based on limited data.

UK Forecast

Party % Vote Forecast Change on 2015
 Conservatives 43.4% (41.0% – 45.8%) 343 – 351 +13 to +21
 Labour 36.0% (33.5% – 38.5%) 219 – 232 -12 to +1
 SNP 3.9% 43 – 53 -13 to -3
Liberal Democrats 7.4% (6.6% – 8.2%) 5 – 7 -4 to -2
Plaid Cymru 0.7% 3 – 5 same to +2
Green 1.8% 0 -1 -1 to same
UKIP 4.2% (3.5% – 4.9%) 0 -1
Speaker 1
Northern Irish 18

Most likely result – Conservative Majority of 48

We have specific Welsh and Scottish forecasts as well.

Scotland

Party % Vote Forecast Change on 2015
SNP 41.7% (37.6% – 45.8%) 43 – 53 13 to -3
Conservatives 27.6% (23.0% – 31.7%) 5 – 10 +4 to +9
Labour 23.7% (18.2% – 29.2%) 0 – 4 -1 to +3
Liberal Democrats 6.1% (5.3% – 6.9%) 0 – 3 -1 to +2
Greens 0.4%
UKIP 0.4%

Probability of Conservatives in second place : 62.2%

Wales

Party % Vote Forecast Changes on 2015
Conservatives 39.0% (29.5% – 48.5%) 16 – 21 +5 to +10
Labour 33.3% (26.0% – 40.3%) 14 – 19 -11 to -6
Plaid Cymru 14.8% (11.5% – 18.1%) 3 – 5 same to +2
Liberal Democrats 5.3% (2.3% – 8.3%) 0 – 2 -1 to +1
UKIP 5.3%
Greens 0.7%

Probability of Conservatives in second place : 60.9%

London

Party % Vote Forecast Changes on 2015
Labour 49.1% (45.2% – 53.0%) 41 – 47 -4 to +2
Conservatives 38.0% (32.5% – 43.5%) 26 – 32 same to +6
Liberal Democrats 7.5% (4.2% – 10.8%) 0 – 1 -2 to -1
UKIP 2.9% (0.6% – 5.2%) 0 same
Greens 2.4% (1.4% – 3.4%)  0  same

Northern Ireland

Party % Vote Forecast Changes on 2015
DUP 28.5% 7 – 12 -1 to +4
Sinn Fein 27.0% 3 – 6 -1 to +2
SDLP 12.4% 1 – 3 -1 to +1
UUP 14.8% 0 – 2 -2 to same
Alliance 9.8% 0 – 1 same to +1
Independent 1 same
Greens 1.0%

Leave a Reply