New Times,
New Thinking.

  1. Politics
  2. Polling
4 January 2024

To win, the Tories need to defy electoral history

No governing party has ever won from a starting position as weak as the Conservatives’ currently is.

By Ben Walker

Ahead of what Rishi Sunak has confirmed will be a general election year (the Prime Minister ruled out a contest in January 2025), Labour is in a better situation than any opposition in recent history to replace the government.

Keir Starmer’s party stands 19 points clear of the Conservatives. This lead is unlikely to be overturned, but that isn’t to say it won’t shrink in advance of election day. History tells us that they tend to narrow. In 1979, for instance, the Tories’ lead over the then Labour government fell from 20 points 90 days before to seven points on election eve. In 1970, Ted Heath’s Conservatives led by 15 points 100 days before, but ended up winning by just three.

In 1997, while Labour’s huge poll advantage of around 20 points barely dipped in 1996-97, the election result was closer: the party finished just 13 clear. The general pro-Labour bias in the polls could be attributed to the party’s high approval ratings under Tony Blair, which discouraged more sceptical voters from engaging with the democratic process altogether. This partly explains the overall fall in turnout, both in 1997 and 2001.


The most recent change elections I’ve looked at (2010, 1997, 1979 and 1970) have seen the opposition’s poll lead end up as much as three quarters the size of the lead 120 days out (1979), or a mere fifth of it (1970). If you apply that kind of range to Labour’s current lead, the party’s advantage over the Tories after the next election could be as high as 14 points or as low as four. That’s what we call poll drift.

How useful are comparisons with past elections? Firstly, polling in sample terms is much improved and has become more accurate with every decade that has passed (despite some notable misses). Secondly, there are alternative metrics that can tell us who’s likely to win the election and by how much. Unlike Ed Miliband and Jeremy Corbyn before him, Keir Starmer leads the prime minister of the day for likeability by 38 per cent to 26 per cent in a head-to-head contest.

[See also: Has Labour dropped its £28bn green pledge?]


Yet when compared with past change elections, brand Starmer still appears a weak spot for Labour. Thirty-eight per cent of voters like him but 37 per cent do not. Contrast this with David Cameron or Tony Blair, who both enjoyed the approval of 50 per cent of the public, and you could be forgiven for believing, in isolation, that Starmer is on course for defeat.

Give a gift subscription to the New Statesman this Christmas from just £49

But Labour’s anticipated victory this year is unlikely to be built or broken on leadership alone. On perhaps the most crucial metric for gauging a party’s competence – whether it is trusted to run the economy – Starmer’s party is ahead, and by almost as much as in 1997.


In January 1997, more than 100 days from the election, Labour led the Tories by 13 points on the economy (42 per cent to 29 per cent). In January 2024, Labour leads the Conservatives by 11 points (39 per cent to 28 per cent).

These numbers do not suggest anything other than a significant Labour victory. It would be foolish to dismiss the election as a foregone conclusion in a volatile age but it would also be foolish to suggest an alternative outcome is likely.


Past results suggest that large opposition poll leads narrow ahead of elections; Labour is unlikely to win by 18 points. But the record also shows that no governing party has ever won from a position as weak as the Conservatives’ is currently. To secure a fifth consecutive victory, the Tories will need to defy history.

[See also: Why Rishi Sunak has ruled out a May general election?]

Content from our partners
When partnerships pay off
Breaking down barriers for the next generation
How to tackle economic inactivity

Topics in this article : , ,