Reason to Doubt Uxbridge Constituency Poll's Boris Bounce
New constituency polling, which suggests Boris would win an election in Uxbridge and South Ruislip, has been doing the rounds on social media this morning and it's easy to see why. The headline finding is that Boris would secure support from a majority (50%) of Uxbridge voters, with his nearest challenger, Labour's Danny Beales, on 33%. Based on events of the last few years, and the current state of national polling, it's clearly a sensational result.
As with most counter-intuitive individual poll results, it should be treated with caution. Even more-so considering the chequered history of constituency polling. In the 2019 election, Boris Johnson achieved 52.6% if the vote, so his vote share has declined in this poll, however Labour's vote share has declined by more (4.6%). This leaves the poll with a 1% swing from Labour to the Conservatives. In the context of a national figure closer to -13.7%, it seems hard to believe.
In his piece on Conservative Home, Lord Ashcroft argues the result should be taken at face value because "Johnson is able to command extraordinary affection and loyalty". Whilst this may well be true, and there is undoubtedly some advantage to his incumbency and near unparalleled name recognition, there is little reason to suppose this advantage would have increased since 2019. At the last general election Boris was standing in Uxbridge as the incumbent Prime Minister leading a national campaign, it's highly unlikely his personal appeal has improved since then.
Taking a look into the tables perhaps provides a better explanation for the result. Interviews for the poll were conducted via telephone, which has a well-documented tendency to skew towards certain demographics - as represented in this poll which oversamples older voters and women. Whilst this sampling bias is then weighted for, some of the figures in this poll - namely for young voters - seem dubious.
According to the poll, 18-24 year olds in Uxbridge would vote for Boris Johnson over Danny Beales by a margin of 9% - this was only based on a subsample of 52 and as such was weighted for three times over. This group be contrasted to 18-24 year-olds nationally, who prefer Labour to the Conservatives by a margin of around 50%. The same pattern, although of a smaller magnitude, is also present in the 25-34 subsample.
Although it's by no means an exact science, and doesn't account for Boris's popularity, applying an estimate for the generic voting intention based on YouGov's latest tables to these two subsamples alone would shift the overall poll result to put both candidates on 42% - with Danny Beales narrowly winning out. The poll should be treated with caution.
