Our new Voting Intention tracker finds the Conservatives as the most popular party with Labour pushed into third, polling 1 point lower than Reform UK

Why is the Labour vote share so low?

The Labour vote share in this poll is almost a third lower than their 2024 GE total. There are several explanations for this:

• Labour voters are less likely to say they’ll vote. Only 84% of 2024 GE Labour voters say they would ‘definitely’ or ‘very likely’ vote if there were a general election called tomorrow, compared to 95% of Conservatives and 90% of Reform UK voters
• Labour voters are less likely to stick with their party. Of those who say they’re planning to vote and give a valid vote choice, only 69% of 2024 GE Labour voters say they’ll vote Labour again, compared to 82% of Conservatives and 90% of Reform UK voters
• Labour voters are more likely to say they ‘Don’t know’ how they would vote. 2024 GE Labour voters are more than twice as likely as Conservative and Reform voters to say ‘Don’t know’, even after prompting them with a follow-up question

Notably, this Find Out Now voting intention also records a lower Labour vote share than other polling companies. A key reason why is the voting intention methodology used:

Firstly, unlike some polling companies which have recently produced UK voting intentions (such as Techne or Deltapoll) we explicitly ask about people’s likelihood to vote – before asking them which party they plan to vote for – which factors into our final headline VI calculation. In our experience, running a voting intention as a single question with a ‘would not vote’ option encourages respondents to select a party even if they are unlikely to vote, which inflates declared turnout. Given Labour voters are less likely to say they’re planning to vote, not accounting for this understates the impact of turnout differentials in the headline calculations.

Secondly, when respondents say they ‘Don’t know’ how they plan to vote (even after prompting) we exclude them from the final headline VI calculation. This approach differs from other companies. Some pollsters (such as Opinium and More in Common) impute the voting intention of these respondents based on other respondents in the same poll with similar demographics and previous voting behaviour. Others (such as BMG) explicitly reallocate these respondents based on how they previously voted – i.e. a 2024 Labour voter who now says they ‘don’t know’ how they would vote, is counted towards the Labour total.

While those approaches allow for a more stable voting intention series over time, our approach makes our voting intentions more sensitive to the current stated voting behaviour of respondents. Both approaches are valid – but helps explains why our final headline vote shares are different to other pollsters.

Our methodology:

• We ask respondents how likely they would vote if there were a general election tomorrow. To those who said ‘definitely’ or ‘very likely’ we then ask their vote choice
• If a respondent answers ‘Don’t know’ they see an additional question, prompting them again to give an answer. If they then give a valid vote choice they are included in the final calculation, but if they answer ‘Don’t know’ again they are excluded from the final headline calculation
• If a respondent was eligible but didn’t vote at the last general election, they are only included in the final headline calculation if they say they are ‘definitely’ going to vote this time. This is because past voting behaviour is generally more predictive than their declared intentions
• The final sample is filtered and weighted (more details here) to be nationally representative by Gender, Age, Region and 2024 General Election vote
• A deeper discussion of our methodology in the context of the 2024 General Election can be found here

Full data tables can be found here