Election Day forecast

Forecast main 150507

Good morning! As the first votes are cast, our final forecast makes Ed Miliband the slight favourite to lead the next government, with a 58% chance to David Cameron’s 42%.

There’s no change in our polling forecast: the Tories still lead by 1 point, 34%–33%, and we make them clear favourites to win the most votes, with a 77% chance.

However, we’ve identified more constituency-level effects that change how our model expects those votes to translate into seats. In particular, the Tories are doing less well to UKIP on the east coast, and to the Lib Dems in LD/Con incumbent seats. Meanwhile, the Lib Dems are suffering more and Labour gaining more in LD/Lab seats, so there’s little net change for the Lib Dems. The overall effect is to move our seats forecast slightly away from the Tories and towards Labour.

As a result, the Tories’ chances of winning the most seats are down to 68%, and their chances of a majority have fallen to just 6%. (Labour’s haven’t improved much though. They’re still less than 0.5%.) A hung parliament looks more likely than ever, with a 94% chance.

Our central forecast is now for the Tories to win 285 seats, leaving them 38 short of a majority. Labour are 23 behind, on 262, while the SNP get 53 and the Lib Dems 25. That would leave the Tories unable to form a majority grouping (even Con+LD+DUP+UKIP would be one short, on 322), and Labour reliant on the SNP and either the Lib Dems or all the other left-wing MPs.

We’ll find out soon!


Date of forecast: 7 May 2015
Days till the election: It’s today! (Go vote!)

Inputted current average poll shares
Con: 34%
Lab: 33%
LD: 9%
UKIP: 13%
Others: 11%

Forecast GB Vote Shares (with 95% Prediction Intervals)
Con: 35% (31% – 39%)
Lab: 32% (28% – 36%)
LD: 10% (7% – 14%)
UKIP: 12% (8% – 16%)
Others: 11% (9% – 12%)

Forecast Scotland Vote Shares (with 95% Prediction Intervals)
SNP: 48% (44% – 52%)
Labour: 27% (23% – 31%)

Forecast GB Seats (with 95% Prediction Intervals)
Con: 285 (245 – 326)
Lab: 262 (223 – 300)
LD: 25 (17 – 33)
SNP: 53 (45 – 57)
PC: 3 (2 – 3)
UKIP: 3 (3 – 4)
Grn: 1
Other: 1
(May not sum to 632 due to rounding of sums of probabilities.)

Central forecast: Con short of a majority by 38
(Criterion for majority now changed to 323 not 326, assuming Sinn Fein win 5 seats and do not take them.)

Probabilities of key outcomes
Con most votes: 77%
Lab most votes: 23%

Con most seats: 68%
Lab most seats: 32%

Hung Parliament: 94%
… with Con largest: 62%
… with Lab largest: 32%

Probabilities of predicted government outcomes:
(See here for explanations and assumptions)

Con majority: 6%
Con+DUP: 5%
Con+LD: 16%
Con+LD+DUP: 14%
Con largest, Lab+SNP+DUP+other left maj: 5%
Con largest, but Lab+SNP+other left maj: 15%
Con largest, but Lab+SNP maj: 6%
Lab largest, Lab+SNP maj: 8%
Lab+LD+DUP+SDLP+PC+Grn+Hermon: 9%
Lab+LD+DUP: 7%
Lab+LD: 7%
Lab+DUP: 1%
Lab majority: 0%

(Probabilities may not sum due to rounding)

Shares trend

Shares trend b

6 thoughts on “Election Day forecast”

  1. Just to point to small detail regarding MP needed to majority for Con+LD+UKIP+DUP. You mention 322 as short, but I think you are forgeting that the opposition would have the same number, (if SF get 5 seats and abstain), so the Speaker would be called to break the tie and he has to vote for the current Gov’t.

  2. I think, assuming that the exit poll numbers are not unrealistic (and we are assured that the exit poll is normally very accurate!) that we will now be asking one huge question: ‘Can we ever really trust regular (non-exit) opinion polls again? ‘

    The opinion polls before the Scottish referendum last year were quite
    wrong and the polls leading up to the 1992 general election were even more misleading. It now looks like the inaccuracy of the polls at this election will match or beat those in 1992.

    How could so many pollsters assume that Labour could lose around 40 seats in Scotland and still have around 275 seats in the Commons? It never made sense to me – arithmetically or logically……

    Ah well, let’s see how close to 323 seats the Tories can get….

    Perhaps this time the exit poll will prove to be the dud, but I doubt it.

  3. Congratulations to the great British public. They saw what was really at stake, and never became involved in the complex (and in the event, often nonsensical) discussions about improbable coalitions etc. They have delivered a great verdict, and one which I had predicted.

    I found your election posts very interesting, and you actually produced some better forecasts than many of the pundits.

    …….I shall just pop down to the bookmakers and collect my winnings!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s