General Election: Will Keir Starmer beat Tony Blair's record?

• Labour task is much harder than 1997

• Biggest progress coming in unlikely areas

• Starmer's Labour rebrand working for now

• Read our UK General Eleciton live blog here

In the wake of more terrible polls for the Conservatives, Betfair markets signal that indeed, Keir Starmer is on course to beat Labour's all-time record tally of 418 parliamentary seats, set in 1997 under Tony Blair. Midweek, 'Yes' in our Over 418.5 Labour Seats was trading at 2.226/5. Now it is odds-on at 1.845/6. Still, nevertheless, a close betting heat.

Polling models project historic total

There are also plenty of signals from polling models. A new projection from FindOut Now/ElectoralCalculus on Friday estimated that Labour will smash that record by a sizeable margin, winning 476 seats. Their MRPs have consistently shown this over the past 18 months, having previously estimated 452, 461 and a quite extraordinary 509 back in January 2023.

Whilst their surveys have produced the highest returns for Labour, they aren't alone. During the same 18 month period, Survation produced MRPs estimating 468, 426 and 475. However the last two MRPs from Yougov, in March and December, produced notably lower numbers, 403 and 385. A couple of live polling models fall in between. New Statesman have Labour on 435. Financial Times on 448.

We eagerly await more MRPs during this campaign. Given the polls - even the firm with methodology most favorable to the Tories, Opinium now show a 20% Labour lead - I expect numbers to be on the higher side. Hence the Under 140.5 Conservative Seats predicted in our Live Election Blog last week, and the trajectory across all Betfair seat total markets.

Turnaround much greater than 97

Even the low end of these results imply an extraordinary, historic turnaround. If Blair is the benchmark, Starmer is well ahead. That 1997 result saw Labour gain 146 parliamentary seats, because they started in a much better position compared to 2019. This cycle, they started on just 202 seats, or 200 based on notional results following boundary changes.

Compare the total number of gains to other years when the government has changed. David Cameron gained 96 seats, Margaret Thatcher 62, Ted Heath 77 and Harold Wilson 59 when becoming Prime Minister. The only close comparison is in 1945, when Clement Attlee gained 239 seats, to reach a total of 393.

Labour in unchartered territory

What these numbers show is the scale of the task, and the projected scale of Conservative collapse. Labour are on course to win dozens of seats where they have never done so before. Because, as explained in my piece regarding a Tory wipeout, an unpopular party will struggle to win anywhere under our voting system.

Nevertheless, reaching these extreme targets will require winning in constituencies with no history of voting Labour. And not only extreme targets. For example, based on Labour's target list, they would need to win for the first time in the following places, merely to win an overall majority (92 seats short of Blair's total) - Bournemouth East and West, Monmouthshire, Colchester.

Demographics explain strategy

This to a great extent explains Starmer's strategy since 2019. Labour's image and message has never been suited to constituencies with a low share of public sector workers or students, and higher shares of older voters and those defined as 'white British'. That applies to almost every seat around or below the majority line.

Blair did cut through in 'Middle England'. There are plenty of seats below the majority line that were in the Labour column after 1997. Several Kent seats, Tamworth, Stoke-on-Trent South, Redditch, Amber Valley, Nuneaton, Wellingborough, Tamworth. They gained the last two in recent by-elections.

But under Gordon Brown, Ed Miliband and Jeremy Corbyn, Labour crashed spectacularly in these areas. When Starmer took over, it was unthinkable they could even be competitive in this type of seat.

Purging Left is a calculated gamble

Thus when Starmer promises The Sun that he will cut immigration, this is his thinking. Whilst having any relationship with The Sun sticks in the craw of many Labour voters and particularly members, he calculates doing so is both necessary and advantageous. That there are more votes to be won than lost.

The counter is that he is alienating a large chunk of natural Labour support. See the ongoing rows over the expulsion of Jeremy Corbyn, deselection of Faiza Shaheen and farcical management of the Diane Abbott situation. It could mean losing a safe seat or two. Corbyn certainly has a great chance in his bid to retain Islington North but I'd be highly sceptical either of the other two could win as an Independent, should they try.

This row dominated media coverage last week, giving the Tories a talking point to latch onto, and is still prominent. The general consensus was that it represented, regardless of the individual cases, poor PR on Labour's behalf. But if the weekend polls are a guide, it did no harm.

Proportion of Britons who believe X has had a...

Keir Starmer
Good week: 45%
Bad week: 20%

Rishi Sunak
Good week: 20%
Bad week: 55%

via @OpiniumResearch https://t.co/3KGw5OcJZZ

-- Ben Walker (@BNHWalker) June 1, 2024

Observe recent stunning gains

Changing Labour may have negative long-term consequences. Only time will tell. But for now, it is enabling them to cut through in the unlikeliest of places. Of all their by-election gains, the most stunning by my reckoning was Selby and Ainsty. A rural seat in North Yorkshire, which voted for Brexit.

The constituency has been redrawn for this election but the swing required would have placed it well past the 'Blair line' of 418. Unlike other by-elections, there was no big scandal around the departing Tory MP. Yet they won it easily, by 12%, and are further ahead now in national polls than then.

Also, note recent wins in the mayoral elections in York/North Yorkshire and East Midlands. Labour are winning in places, easily, where they started miles behind. It all suggests these MRP projections and betting signals are accurate. I've backed Over 418.5 repeatedly, most recently at 2.226/5. At 1.845/6, it still looks a decent bet.

Now read General Election: What constitutes a Tory wipeout and how bad could it get?