This week’s estimates from the Ashcroft Model suggest a narrowing of the Conservative majority, though still a comfortable victory for Theresa May. Our “combined probabilistic estimate”, in which we take the sum of each party’s win chances in all the seats in which it is standing, the model gives the Conservatives 355 seats (down from 396 last week), or a potential majority of 60.
However, the majority could be considerably better or worse than this for the Conservatives, depending on the pattern of turnout. Our model calculates three different results, depending on who actually shows up to the polls. If everyone who claims in our surveys to have voted in the EU referendum turns out next week, the number of Conservative seats could fall to 345, with 233 for Labour – an overall majority of 40. But such a surge fails to materialise and turnout is confined to those who actually cast their vote in 2015, our estimated Tory majority rises to 78. Updated vote share and win-chance estimates for each seat are available here.
It is important to remember that we are dealing with probabilities not predictions, and that the result could well fall either side of our central estimates. In the scenario where turnout matches that of 2015, the highest likelihood (35%) is of a Conservative majority between 60 and 79, with a one-in-four chance of a majority between 80 and 99 and a 12% chance of a majority of 100 to 119. But if everyone who says they are certain to vote actually does so, the biggest chance (35%) is of a Tory majority between 40 and 59.
It is worth emphasising how much there remains to play for in the last week of the campaign. According to our 2015 turnout model there are 70 seats where the two leading parties’ estimated vote shares are within 5% of each other. Twenty-three of these tight races are potential Conservative gains, twenty of which would be from Labour, two from the SNP (Moray, and Perth & North Perthshire).
To explore the results by constituency, and for full details of the overall seat estimates, click here.
