The scale of Keir Starmer’s task at the next election is daunting

John Rentoul surveys the size of the electoral mountain Labour must scale to form a government at the next election

The Labour leader during a visit to Sheffield last week (Getty)

Much of the commentary yesterday about Keir Starmer’s first year as Labour leader seemed to take place in a vacuum. Is he doing better than his predecessors, Jeremy Corbyn, Ed Miliband and Gordon Brown? To which the answer is yes, but they all lost. The real question is how well Starmer has to do to win, and on that he remains a long way from the goal.

A statistic sometimes bandied about is that Labour needs to gain 123 MPs to win a majority of one in the House of Commons. This is not quite right. With seven Sinn Fein MPs who do not take their seats, the target comes down to 119 gains.

But then there are likely to be constituency boundary changes by the time of the next election. If the election is in 2023, as Starmer yesterday warned his party it might be, it would probably be after 1 July that year, which is when the new boundaries will take effect. The independent boundary commissions, one each for England, Scotland, Wales and Northern Ireland, have only just begun their work, so it is impossible to predict precisely what effect the new boundaries will have. However, the boundaries have been biased against the Conservatives for a long time – Tory MPs tend to represent constituencies with larger populations than Labour MPs – so any attempt to equalise constituency sizes is going to favour the government.

The last attempt to redraw boundaries was abandoned because David Cameron sought at the same time to cut the number of MPs from 650 to 600, which meant that too many Tory MPs feared losing their jobs. But those new boundaries would have had the effect of transferring about 16 seats from Labour to Tory. That figure may be lower in the current boundary review because the Tories won so many traditional Labour seats, with their lower populations, at the last election. But if we assume that the boundary changes are worth 16 seats, that means Labour’s target will be to gain 135 seats.

That is an immense challenge. It would require a swing in votes of 12.5 per cent, which is higher than that achieved by any party since the war – the biggest swing so far was 10.2 per cent to Labour under Tony Blair in 1997.

Swing is the best measure of the change in a party’s performance from one election to the next, because it measures the change in the gap between the share of the vote won by each of the two leading parties, which is what decides elections. Jeremy Corbyn’s supporters are fond of selectively citing statistics about the 2017 election, in which Labour’s vote share went up by 9.8 percentage points, ignoring the increase in the Tory vote share of 5.7 points. That means that Labour closed the gap by 4.1 points, which is the equivalent of 2.0 per cent of voters (after rounding to one decimal place) “swinging” from Tory to Labour. So the swing in 2017 was 2.0 per cent, a pretty anaemic result by postwar standards (and more than wiped out by the 4.5 per cent swing to the Tories in 2019).

So Starmer would have to do better than Blair, and more than six times better than Corbyn in 2017, to scrape a bare majority in parliament.

If that seems unrealistic, what would it take for Starmer to be able to form a minority Labour government in a hung parliament? If we assume that the only party with MPs in the Commons that would prop up a Conservative government is the Democratic Unionist Party (DUP), as they supported Theresa May in 2017, and that they retain their current eight MPs, then Starmer’s target is to reduce the number of Tory seats to 314. That would give all the anti-Conservative parties together a majority of one.

Of course, there would then be the complication of having to deal with the Scottish National Party, the Liberal Democrats and others – although they may not be as great as some in Labour might fear, because none of the opposition parties apart from the DUP would contemplate allowing the Tories to stay in office.

Again assuming new boundaries give the Tories 16 extra seats, this hung parliament scenario would require a swing of 5 per cent, which is still a big challenge, if not an unprecedented one. There have been only three elections since the war when the swing was as much as 5 per cent: Margaret Thatcher in 1979 (5.3 per cent), Blair in 1997 (10.2) and Cameron in 2010 (5.1).

Even to get to the point where Starmer faces the problem of negotiating with the SNP requires one of the biggest shifts in votes since the war.