Keir Starmer would be sitting on a 120-seat majority
A new YouGov MRP model, commissioned by Conservative Britain Alliance and released in The Telegraph last night, shows that Labour would win a large majority were the election to be held today.
The results of the model, which uses data from 14,110 respondents answering between 12 December and 4 January, would see Keir Starmer enter Downing Street having secured a 120-seat majority, with Labour winning 385 Commons seats. This would be a 183-seat increase for the party since the last election.
The Conservatives meanwhile would slump to just 169, losing 196 seats compared to 2019.
These results are reminiscent of the 1997 general election outcome, which saw Tony Blair’s Labour won 418 seats and John Major’s Conservatives took 165.
Meanwhile, the Liberal Democrats would receive 48 seats – also similar to their 1997 haul – making them a notable Parliamentary force once again. The SNP, meanwhile, would fall to 25 seats, with Labour making significant gains in the central belt.
The Greens would hold on to their Brighton Pavilion seat, without making any further gains – although they come incredibly close in Bristol Central at 38% to Labour’s 40%. It is possible that by election day the party will have done enough to convince voters in the area to give them a shot.
Reform UK would fail to win any seats, as the Brexit Party and UKIP did in 2019 and 2017 respectively.
This MRP is our first since new parliamentary constituency boundary proposals have been finalised. While this is the first time an election will be fought using these constituencies, notional results calculations allow us to see what the outcome in each seat would have been if the last election had used those boundaries, and therefore which seats would be changing hands.
Most notably this includes chancellor Jeremy Hunt’s Surrey constituency of Godalming and Ash, which is narrowly won by the Liberal Democrats in this model. Other cabinet casualties include Leader of the House of Commons Penny Mordaunt (Portsmouth North), Welsh secretary David TC Davies (Monmouthshire), attorney general Victoria Prentis (Banbury), and veterans minister Johnny Mercer (Plymouth Moor View).
Conservative party chairman Lee Anderson would also lose his Ashfield seat to Labour, and Boris Johnson’s former Uxbridge constituency would likewise fall to Labour.
Notes on the Daily Telegraph’s analysis
The Daily Telegraph wrote that “In constituencies across England and Wales, the Labour vote is up by an average of just four per cent compared to 2019”. This is somewhat of a red herring. There is a sum using certain notional results whereby the estimated Labour share looks like a mean of a four point rise on their 2019 performance. However, this is not the correct way to look at either implied national changes nor what is happening at the constituency level.
If we aggregate up all our constituency level figures and then weight them according to likely voter population, the headline vote intention figures come out at the following:
Labour 39.5%, Conservatives 26%, Lib Dems 12.5%, Reform 9%, Greens 7.5%, SNP 3%, Plaid 0.5%, Others 2%.
A separate note by the Daily Telegraph suggested that the presence of Reform UK is the difference between Labour securing a majority and not. This is their own calculation using our data, and appears to be based simply on adding the Conservative and Reform UK vote shares together in each constituency, which is not a reliable way of measuring their impact.
Were Reform UK not to contest the election, it is extremely unlikely that all, or even a majority, of their voters would transfer to the Conservatives. Some would go to UKIP and splinter parties, some to Labour and other established parties, and some would simply stay at home – YouGov polling in October found only 31% of Reform UK voters would be willing to vote Conservative if Reform UK were not standing in their constituency.
Finally, the Daily Telegraph also said that the YouGov MRP model does not account for tactical voting in its estimated shares. This is not the case – our model does provision for tactical voting in its design, including by estimating constituency competition effects as part of the model equation. It does not, however, apply any post-hoc readjustments to vote share estimates based on any assumed model of tactical voting beyond what we already have in the data.
Methodology
YouGov interviewed 14,110 adults across the country between 12th December 2023 and 4th January 2024. Constituency-level forecasts were estimated using the same method which correctly predicted the 2017 and 2019 UK general elections. MRP constituency forecast models first estimate the relationship between a wide variety of characteristics about prospective voters and their opinions – in this case, ‘which party would you vote for if a general election were being held tomorrow' – in a ‘multilevel model’. It then uses data at the constituency level to predict the outcomes of seats based on the concentration of various different types of voters who live there, according to what the multilevel model says about their probability of voting for various parties (‘post-stratification’). In this instance, 500 draws from the posterior distribution of the multilevel model were used to predict the constituency probabilities. The multilevel model was estimated through 12,000 iterations. Turnout likelihood for each voter group was estimated using a multilevel model on British Election Study data. The precise multilevel model equation has been benchmarked to correctly estimate the 2019 General Election to within a couple of seats of each party’s actual performance in that election, and the overall approach to MRP by YouGov has been used to successfully predict elections as recently as Spain in July 2023.
https://news.google.com/rss/articles/CBMif2h0dHBzOi8veW91Z292LmNvLnVrL3BvbGl0aWNzL2FydGljbGVzLzQ4MzcxLXlvdWdvdi1tcnAtc2hvd3MtbGFib3VyLXdvdWxkLXdpbi0xOTk3LXN0eWxlLWxhbmRzbGlkZS1pZi1lbGVjdGlvbi13ZXJlLWhlbGQtdG9kYXnSAQA?oc=5
2024-01-15 12:16:00Z
CBMif2h0dHBzOi8veW91Z292LmNvLnVrL3BvbGl0aWNzL2FydGljbGVzLzQ4MzcxLXlvdWdvdi1tcnAtc2hvd3MtbGFib3VyLXdvdWxkLXdpbi0xOTk3LXN0eWxlLWxhbmRzbGlkZS1pZi1lbGVjdGlvbi13ZXJlLWhlbGQtdG9kYXnSAQA
