Opinium’s political polling in the 2024 general election campaign

Maintaining accuracy in our political polling

The purpose of our general election polling is to accurately show the levels of support for each party at any given time. In the 2019 General Election Opinium did this to within 1% accuracy for all parties, bar the Greens.

Part of Opinium’s continued accuracy comes from regularly monitoring and updating its methodologies, even when past results have been accurate, to ensure the polling reflects Britain as it is today.

For the 2024 election, here is a summary of what to look out for from Opinium in the final three weeks of the campaign.

PART ONE: Asking voting intention in a general election campaign

1. Tracking voting intention between elections

Between elections Opinium asks voting intention for a general election in two stages.

The first question shows a list of major parties standing across the UK and an option to tick ‘other’ Those who tick ‘other’ are taken to a secondary question where they are shown another list of smaller parties, determined by a number of factors including the coverage of the parties in nationwide elections such as the number of candidates standing in local elections.

It is well-documented that showing all parties together consistently overstates support for minor parties when these are tested in actual elections. This was seen in the 2015 and 2017 general elections, and the 2019 European Parliament elections, the analysis from which helped Opinium design its accurate methodology for the 2019 general election.

2.Tracking voting intention during a General Election

During a general election, the two-stage approach changes.

The main change is that as well as asking the usual two-stage voting intention question, participants now see a third question showing the parties standing in their constituency and asking them who they will vote for.

This is a change that can only take place after nominations have closed and there is a definitive list of which parties are standing in which constituencies. In 2019, for example, the effect was that around 40% of people were no longer shown the Brexit Party because they lived in constituencies the Brexit Party was not standing in.

On Friday 7th June 2024, nominations closed for candidates standing in the 2024 general election. Therefore, Opinium can now make final adjustments to its questions about the actual election taking place on 4th July 2024, rather than a hypothetical “general election held tomorrow”.

3. How will Opinium report headline voting intention in 2024 General Election

Therefore, from week commencing 10th June 2024, Opinium will report voting intention figures based on the constituency question rather than the usual two-stage approach. For completeness, Opinium will still ask the two-stage questions as well, so that it can tell how much of a change in each party’s vote share is down to the way the question is asked. Maintaining both questions allows Opinium to judge which poll movements are due to this change rather than a shift in actual opinion.

The first poll with this change will be published at 8pm on Saturday 15th June 2024. Opinium’s full schedule of polls for the rest of this general election campaign are as follows:

Fieldwork periodReleaseNotes
12th to 14th June 20248pm Saturday 15th June 2024 
19th to 21st June 20248pm Saturday 22nd June 2024 
26th to 28th June 20248pm Saturday 29th June 2024 
1st to 3rd July 2024Early afternoon, Wednesday 3rd July 2024This will be our final poll of the campaign.

PART TWO: Ensuring Opinium’s sample composition is up to date

  1. How Opinium gets its weighting targets and ensures its sample is representative

Like the rest of the polling industry, Opinium  relies heavily on the statistics authorities, such as the ONS, across the UK provide, for the most up-to-date weighting targets possible.

As a reminder, the factors Opinium weight are:

  • Gender, age, and education levels interlinked
  • Region
  • Ethnicity
  • Working status
  • Political attention levels
  • 2019 past vote
  • EU Ref vote

2. How Opinium takes account of past voting behaviour

Best practice for polling firms is to ensure that that they are making their sample as politically representative, as well as demographically representative, as possible in the run up to a major national election. For Opinium, this involves updating the targets for two key criteria.

Political attention levels change during an election campaign. For this reason, Opinium has  to account for a slightly larger proportion of people who say they are highly politically engaged. Opinium also combines ‘low’ and ‘mid’ levels of political attention, because, during an election campaign the number of people with very low political attention drops significantly due to politics consistently being the main news event.

Opinium has also adjusted the past vote targets to account for the different population makeup over a five-year parliament. This is the longest gap between general elections since the 2010-2015 parliament, and therefore targets have been adjusted to account for how the population composition has changed in that time.

3. How Opinium weights headline voting intention

Opinium will continue to weight its voting intention question differently to the rest of its political polling questions. Firstly, this will only take those respondents who give a voting intention (e.g. removing those who are undecided or say they won’t vote). Secondly, this will ensure those are left are representative of those who have voted in recent general elections.

This is similar to the approach taken by MRP models that have accurately forecasted the results of the last two elections. 

These targets have been developed by looking at past turnout data from the British Election Studies and is how Opinium will continue to weight its headline voting intention.

PART THREE: Opinium’s track record and where to go for more information

Opinium is proud of its track record in accurately building representative samples of the British public.

Opinium was the most accurate research agency in both the 2019 UK General Election and the 2016 EU Referendum.

For those interested in the details, we have provided two key resources from our archives outlining how we have conducted our polling over the past five years:

  • A review of how Opinium conducted its polling at the 2019 general election including the additional stages added during the general election campaign. (Link here).
  • Opinium’s mid-term review of sampling and weighting targets that ensure it upholds commitments to the principles of the Sturgis review, and review of Opinium’s final poll in 2019 which accurately forecast the general election result. (Link here).

If you are interested in learning more in the run-up to the 2024 general election, sign up to Opinium’s Substack to keep up to date.