Britain’s general election—which took place on July 4th—brought just the third change of governing party since 1979. As our poll tracker demonstrates, support has ebbed away from the Conservative Party in recent years to the benefit of the opposition Labour Party.
But who is driving these political shifts? At almost every British general election, politicians and the press find a new stereotypical voter to explain the outcome. Think “Mondeo man” or “Worcester woman” (it helps if these groups are alliterative). The Economist wanted to take a more granular look. We have built a model of individual voting behaviour, based on the views of nearly 100,000 Britons, that allows you to build your own British voter. Using data from WeThink, a polling firm that conducts a weekly survey of voting intention, our model lets you explore 905,520 groups that make up Britain’s electorate, defined according to eight demographic characteristics (region, age, sex, ethnicity, education, employment, housing and population density).
Use the tool below to build your own voter, or press shuffle to select one at random, and see how they are likely to vote. The first chart shows how each characteristic shifts an individual’s probable vote from the national average; the second shows how this voter compares with other groups of voters, and how many Britons share the same characteristics.
Build a British voter
Chance of voting for party*, %
undefined
undefined
undefined
undefined
undefined
Other
How this group compares to others
White women aged 30-44 from London, who live in a mortgaged homeundefined, are students and were educated to A-level
*Excluding don't know or won't vote
The model allows us to go beyond the headline polling numbers to explore the dynamics of the electorate. The profile below—a white man between the age of 55-64 living in the north-west of England—shows the type of person who is most likely to have abandoned the Conservatives.
Biggest swing away from Tories
White men aged 55-64, who live in a mortgaged home in a rural area of the north-west, who have no formal qualifications and work part-time
There are about 600 people like this. In 2019 this cohort, and those similar to it, helped Boris Johnson break through Labour’s “red wall”; according to our model, this voter was estimated to have an 83% chance of voting Conservative that year. If the election were held tomorrow, their chances of voting Conservative would be a fraction of that. Both Labour and Reform UK, a right-wing party, stand to benefit. Such stark shifts have a strongly regional character: someone with the same demographic profile in the south-east of England has around a one-in-three chance of supporting the Tories, though they are still most likely to vote Labour. As our article on Britain's changing electorate explains, the tide is turning against the Conservatives everywhere. But only by building more detailed profiles of voters can you see the factors that help explain their preferences. ■
Methodology
Our model is based on survey data provided by WeThink, which obtains responses from a nationally representative sample of approximately 1,250 Britons each week. We gathered all results from September 23rd 2022 to April 25th 2023, amounting to nearly 100,000 individual responses. We have removed people who said that they “will not vote” or “don't know” which party they would vote for if the general election were held immediately.
To estimate voting intentions based on demographic profiles, we fit a multinomial logistic regression model using the LASSO method, a statistical technique that eliminates or reduces the impact of certain variables in order to maximise accuracy on unseen data. Our model accounts not just for the eight demographic features detailed above in isolation, but also for how they interact with each other. For example, switching the listed age group from 75+ to 18-24 sharply increases the chances that a white British voter will support Labour, but has little effect on this probability for someone of Pakistani ethnicity.
Our model also incorporates the national poll average for each party in each week. As a result, if a party gains or loses ground in the polls overall, the model will automatically shift that party’s vote-intention probabilities for each demographic profile in the same direction. We estimated population figures using the 2021 census for England and Wales, and the 2011 census for Scotland. We update the model every week to account for additional survey data and new national polling averages.