The Sunday Times asked me to write up the results of Survation’s latest seat-by-seat projection for the general election. It was commissioned by the campaign group Best for Britain and shared with the Sunday Times in advance.
I was critical of Survation’s last MRP results in April. In particular, they showed the Liberal Democrats doing implausibly badly in many of their target seats and ending up with only 22 MPs. The latest survey seems to have fixed this. It still shows unlikely swings to the Conservatives in the safest Labour seats (including Keir Starmer’s), but this has no effect on the overall tally of seats.
It is worth noting that the overall Labour lead in this latest survey is 16 per cent – historically high, but at the low end of recent polls. In my judgement this is realistic.
Labour is on course for a 262 seat majority in next month’s election, according to a seat-by-seat analysis by Survation. This would be by far the biggest majority won by any party since the second world war. The Conservatives’ projected tally would be just 72 seats. It would be less than half their previous worst election since their formation nearly two centuries ago (156 seats in 1906).
However, the survey, commissioned by Best for Britain, still shows the Conservatives as the main opposition party – as long as they don’t slip further. On Survation’s figures Labour would need a further swing of only two per cent to capture another 43 Conservative seats, and reduce their tally to just 29.
However, in that case, the main opposition party would not be Reform. They would have just seven MPs even though they would come third in votes, with 12 per cent. The Liberal Democrats, with 11 per cent support, would win 56 seats. Ed Davey would become leader of the opposition.
Overall, Labour is set to win 70 per cent of the seats in the new parliament with just 40 per cent of the total vote. The other parties will have only 30 per cent of MPs, despite gathering 60 per cent of the vote. (These figures exclude Northern Ireland, with its different party system.)
On these figures, Labour is set win a host of seats for the first time, such as Chipping Barnet, Hexham, Sutton Coldfield and Tatton. Southend, Bournemouth and the Isle of Wight ever never had a Labour MPs. Now the three could have two each. If Keir Starmer has a constituency problem in Downing Street, he will be the first prime minister to have a local Labour MP to call on.
Survation polled 22,000 respondents, partly by telephone, partly online, between May 31 and last Friday. This huge number means that different groups of voters can be looked at in great detail – by region, age, gender, education, employment, past vote and so on. The detailed data are then matched to the profile of each constituency. This method of converting national vote shares into seats won and lose has attracted some controversy. But the 21-point collapse in the Conservatives’ share of the vote, from 45 per cent in 2019 to 24 per cent in this survey, breaks all recods. No election in modern times offers a precedent for us to draw on.
Moreover, all such MRP surveys are subject to a margin of error when considering purely local factors, such as the reputation of the sitting MP or local controversies over, say, planning or the future of a local hospital. Nevertheless, there is enough evidence from by-elections, local elections and other polls to confirm that the Conservatives are in for an unprecedented battering in even traditionally safe seats. If they slip further, they can no longer be certain to survive as Britain’s second largest party.
For example, Jeremy Hunt seems to be in trouble in Godalming and Ash where Survation reckons that Paul Follows, the Liberal Democrat candidate, will have 44 per cent of the vote,16 points ahead of the Chancellor’s predicted 28 per cent. Local factors seldom overcome that kind of deficit.
Reform’s leader, Nigel Farage is projected to win Clacton narrowly, by 31-29 per cent over Giles Watling, the defending Conservative MP. However, this is an example of where an obvious local factor – Farage’s personal profile – could well enhance his vote and his majority.
However, local factors cannot disguise the fact that Britain’s electoral system can produce apparently weird outcomes. The explanation is that it has three particular properties, all of which are more potent than usual in the current election. The first is that it rewards successful parties. No majority party since the second world war has reached 50 per cent of the vote. More people have always voted against it than for it. In 2005, Labour secured a comfortable majority of 66 seats on just 36 per cent of the bote across Britain.
Second the “success bonus” is enhanced by a big lead over the main opposition party. Tony Blair’s landslide victory in 1997 was achieved with 44 per cent of the vote. This was almost exactly the same the same percentage of the vote as his party obtain in 1964, when Harold Wilson scraped home with a majority of four. But whereas Labour defeated the Tories by 13 per cent of the vote in 1997, its lead in 1964 was just two percent. Labour’s current support is less than in both those elections – but its lead is bigger.
Thirdly, our voting system punishes medium-sized parties with evenly-spread support. The SNP benefits by fighting in only one part of Britain. As for the Lib Dems, Survation projects that they could fall short of their total vote last time – but because their support is more concentrated than before in seats they can win, they are on course to have five times as many MPs.
In contrast, Reform has not (yet?) built up enough support in enough constituencies to win more than a handful of seats. It would take a huge jump in their overall support in the remaining weeks of the campaign to fill more than a single bench in the House of Commons.
Survation’s figures bring good news to the SNP, which is projected to remain the country’s largest party, with 37 MPs (down 11 since 2019) compared with Labour’s 14 (up 13). The Conservatives one (down five) and Lib Dems 5 (up 3). (These figures take account of the new boundaries.)
This outcome is far better for the SNP than recent polls indicate. This is not because of the MRP methodology, simply that Survation’s poll is the first by any company since February to show the SNP ahead of Labour in votes, by 37-30 per cent. In less than three weeks we shall know if that is right.
Peter Kellner. The Politics Counter 