Starmer's biggest mistake is taking progressive voters for granted.

New research by YouGov shows that, even with its support down to 23 per cent, Labour may not have reached rock bottom

I have bad news for Labour supporters. Its problems are even worse than I thought.

Five weeks ago I reported that Labour was losing more votes to the Liberal Democrats and Greens than to Reform and the Conservatives. It was the biggest reason why Labour has lost one third of its support since last year’s general election – down from 35 per cent across Britain then to just 23 per cent in YouGov’s latest poll.

Fresh data show that the haemorrhage may not be over. From time to time, YouGov asks a set of extra questions about party loyalties. It asks respondents to say how likely they are to consider voting for each national party. They offer a scale, from 0 (“I would never consider voting from them”) to 10 (“I would definitely consider voting them”). These shows us how loyal they are to their current party – and which they might support if they change their minds.

To avoid us all being swamped by too many numbers, I have combined responses into three groups: “hostile” (0-3), “sympathetic” (7-10) and neither (4-6).

Let’s start with the overall picture. This is how Britain’s electorate views each party:

Consider first the “sympathetic” column. The figures for all the parties are similar, and very low.

At first sight Reform’s figure seems to be less than their 28 per cent vote share. But voting intention figures exclude those who say don’t know, wouldn’t vote or refuse to say. For comparison with the table above, the voting intentions of the whole sample are: Reform 20 per cent, Labour 17, Conservative 12, Lib Dem 11, Green 7, others 4, don’t know etc 28.

The real difference between the parties is that Reform polarises voters more than any other party, with the largest figure for hostility and the smallest number for neither.

The story was different a year ago. On July 1, at the start of election week, Labour’s figures were: sympathetic 35 per cent; hostile 42 per cent, net score minus seven. In contrast, the Tory figures were: sympathetic 19 per cent, hostile 62 per cent, net score minus 43. Since then, the two parties have converged. The Conservative score has improved fractionally, while Labour has slumped.

Those figures provide the starting point for looking at Labour’s woes. Here are YouGov’s latest figures for those who voted Labour last year

Fewer than two-thirds of Labour voters are sympathetic to the party today. Fifteen per cent – almost 1.5 million – are hostile. Significantly, few Labour voters have any time for Reform or the Conservatives. Those who are looking around are overwhelmingly eyeing the Lib Dems and/or the Greens.

Those figures confirm that Labour has been losing more support to other progressive parties than to the Right. But what about Labour’s current supporters: at 23 per cent support has it hit rock-bottom? Can it rely on those who still say they would vote Labour after tall the decisions, set-backs and bad headlines of recent months?

The short answer is no. This is why I fear Labour’s problems are worse than I thought:

Eighty-one per cent of current Labour supporters tick one of the 7-10 boxes for considering whether to vote for their party. Not bad but far from perfect. And the tiny numbers sympathetic to Reform or the Conservatives bring some comfort.

However, the figures for those voters who are considering voting Lib Dem and/or Green should terrify Keir Starmer. This early in the Parliament it is plainly absurd to predict that Labour won’t recover from today’s catastrophically low support. But it is equally absurd to assert that the party can go no lower. It could easily shed more votes to the Lib Dems and Greens (and possibly the SNP in Scotland and Plaid Cymru in Wales) than it has already.

Why, then, is Starmer so much keener to head off the less threat from Reform than the greater threat from rival progressive parties? His allies make two points. The first is that Reform came second last year in far more seats (89) than the Greens (39) or Lib Dems (6). This is true but not the point. As I reported last week, Labour is defending majorities of up to ten per cent in 104 constituencies. These are its critical marginals. Lose them and it loses its Commons majority. In no fewer than 85 of them the Tories are in second place. Labour defections in these seats to the Greens and Lib Dems could hand local victory to the Conservatives and leave Labour in trouble at Westminster.

This leads to the second argument that Starmer’s allies make. In the seats that matter, progressive voters will not want a Tory or Reform MP. Last year, we are reminded, they voted tactically to kick out the Conservatives; they will do the same next time to keep the Tories out.

If only winning was so easy. Sadly, from time to time, number crunchers must step in to puncture the comfort of simple assertions. This is such a time. The tactical voting argument is not totally wrong; but the numbers show that it’s no get-out-of-jail card.

Here is what happened last year. Labour gained 182 seats from the Conservatives. It dominated the anti-Tory vote but did not monopolise it. Labour’s average vote in these seats was 17,601, while Liberal Democrats averaged 2,871 and the Greens 2,616. Labour won just over three-quarters of the three-party progressive vote, while the Lib Dems and Greens won almost a quarter between them.

So, yes, a fair number of progressives did vote tactically, but quite a few did not. And this was an election ideally suited to tactical voting. The Tories were hated, and Labour said little to put progressives off.

Since Starmer became Prime Minister and Racher Reeves went to the Treasury, they have restricted the winter fuel allowance, cut overseas aid, diluted their climate change policies, reduced disability benefits, prepared public spending cuts, moved right on immigration, and offered Donald Trump a state visit. Dislike of these policies extends well beyond the Corbynite left. (We shall discover in due course how pro-European voters judge the government’s deals with the EU.)

Even so, surely progressive voters won’t want a Conservative or Reform MP? Maybe. But can Labour rely on them to hold their nose and vote for a government that they feel has let them down? Some will, some won’t. That’s why it comes down to the numbers – and why YouGov’s latest findings matter. Even with Labour bumping along on 23 per cent, it may not have reached its irreducible core vote. A significant portion of its remaining supporters might consider switching to the Lib Dems or Greens.

Perhaps the government will do so well over the next four years that Labour won’t need the tactical support of today’s disappointed progressives. It is more likely that their votes will be vital. Are Reeves and Starmer wise to choose policies that put so many of them off?