by Dr Adrian Worton
It's the big day, as the nation goes to vote in an election that really feels like a seismic moment.
We have been tracking the progress of the parties on our General Election page, and here we'll go through our model's final predictions. This article also includes our 'Election dossier', a document to guide you through the big night (see appendix).
It's the big day, as the nation goes to vote in an election that really feels like a seismic moment.
We have been tracking the progress of the parties on our General Election page, and here we'll go through our model's final predictions. This article also includes our 'Election dossier', a document to guide you through the big night (see appendix).
Seat forecasts
First of all, below is our model's 'expected seat count' for each party. This is calculated by summing each party's win probability across all 650 seats. The change in the table below is compared to our article on the 19th of November.
First of all, below is our model's 'expected seat count' for each party. This is calculated by summing each party's win probability across all 650 seats. The change in the table below is compared to our article on the 19th of November.
Party |
Expected seats |
Change |
Conservatives |
344.9 |
+1.9 |
Labour |
217.2 |
+14.4 |
SNP |
42.2 |
-2.2 |
Liberal Democrats |
19.9 |
-13.3 |
DUP |
8.7 |
+0.1 |
Sinn Féin |
6.5 |
+0.1 |
Plaid Cymru |
4.0 |
-0.1 |
SDLP |
1.8 |
-0.1 |
Green Party |
1.3 |
-0.1 |
Brexit Party |
0.7 |
-0.3 |
Alliance |
0.5 |
-0.1 |
UUP |
0.5 |
0.0 |
Change UK |
0.1 |
-0.1 |
Independents/Other |
1.7 |
-0.3 |
Our model, contrary to the notable YouGov MRP model, has the Conservatives slightly extending their majority. The bigger change has been Labour gaining at the expense of the Liberal Democrats, although this really is Labour gaining on the Conservatives, and the Conservatives gaining on the Lib Dems.
We also simulated the model 10,000 times, using the same method as we did in November. Below is each party's range, their 95% confidence interval (where the middle 95% of results fell) and their 50% confidence interval:
We also simulated the model 10,000 times, using the same method as we did in November. Below is each party's range, their 95% confidence interval (where the middle 95% of results fell) and their 50% confidence interval:
Party |
Range |
95% CI |
50% CI |
Conservatives |
251-430 |
280-407 |
320-368 |
Labour |
138-300 |
160-273 |
196-238 |
SNP |
20-57 |
29-52 |
37-46 |
Liberal Democrats |
2-58 |
9-40 |
15-26 |
DUP |
4-12 |
6-10 |
8-9 |
Sinn Féin |
3-10 |
5-8 |
6-7 |
Plaid Cymru |
0-8 |
2-6 |
3-5 |
SDLP |
0-4 |
1-3 |
1-2 |
Green Party |
0-6 |
1-3 |
1-2 |
Brexit Party |
0-11 |
0-5 |
0-2 |
Alliance |
0-4 |
0-2 |
0-1 |
UUP |
0-4 |
0-2 |
0-1 |
Change UK |
0-3 |
0-1 |
0-0 |
Independents/Other |
1-6 |
1-4 |
1-2 |
Interestingly, the overall range for the two main parties has contracted. Amazingly, the model has the Lib Dems occasionally scoring as few as 2 seats. Otherwise, there haven't been any major changes.
As with our November update, we will also look at the number of times our simulations gave the following scenario:
The table below shows the frequency of simulations that gave each scenario:
As with our November update, we will also look at the number of times our simulations gave the following scenario:
- Conservatives have a majority
- Conservative can't form a majority, but they can with help from fellow Leave-ish parties (DUP, Brexit Party, Ulster Unionists).
- Neither the Leave-ish parties nor the Remain-ish parties can form a majority (we are counting Independents and Sinn Féin as neither)
- Labour can't form a majority, but they can with help from fellow Remain-ish parties (all others yet mentioned)
- Labour have a majority
The table below shows the frequency of simulations that gave each scenario:
Scenario |
Frequency |
Change |
Conservative majority |
73.1% |
+3.8% |
Leave-ish majority |
8.8% |
+0.2% |
No majority |
1.1% |
-0.4% |
Remain-ish majority |
16.9% |
-3.8% |
Labour majority |
4-12 |
0% |
Again, we have a contradiction with YouGov, as our model has an increased chance of a Conservative majority. This is despite Labour's gains, and seems to be a result of the Conservatives strengthening their grip on their safe-ish seats.
Conclusion
Labour supporters will take heart from the fact that the predictions aren't as grave as they were in 2017. However, it would be foolish to assume that the polls will all be wrong in the same way they were that time.
The odds are firmly in favour of a Conservative majority, but we'll have to wait and see whether they're right.
Appendix: the TGIAF General Election Dossier
Below you can download our PDF guide to the night ahead:
Conclusion
Labour supporters will take heart from the fact that the predictions aren't as grave as they were in 2017. However, it would be foolish to assume that the polls will all be wrong in the same way they were that time.
The odds are firmly in favour of a Conservative majority, but we'll have to wait and see whether they're right.
Appendix: the TGIAF General Election Dossier
Below you can download our PDF guide to the night ahead:
ge19_dossier.pdf |
As well as an overall summary, the guide has information on all 650 seats, showing the odds in that seat, highlighting any prominent candidates and giving an estimate of when the announcement will be due.