Voters in Stoke-on-Trent and Copeland will turn out on Thursday 23 February to cast their vote in two hotly contested by-elections, where the incumbent Labour party is clinging on.
But they may feel less inclined to do so, if the Met Office is correct. It has predicted that Storm Doris will shake the UK on Thursday, with both Stoke-on-Trent and Copeland in the firing line.
Storm Doris hitting #StokeonTrentCentral hard. Drenched and leaflets like Papier mache. Great team working hard for #Labour4Stoke
— MegHillierMP (@Meg_HillierMP) February 23, 2017
It’s a commonly-held belief that adverse weather can affect the number of voters who make the effort to head to the polls. Ahead of the EU rerendum, the Telegraph ran a news article suggesting that thunder storms on the day could boost the Leave campaign, “which is most likely to benefit from a low turnout”. Labour activists will be particularly nervous in Stoke-on-Trent Central, which has the distinction of being the sole constituency in 2015 where the majority of the electorate did not vote.
Even Anthony Howard once said that evening rain before the polls close “has always been a frightening prospect for Labour”, as working-class voters would often go to the polls between tea and 10pm.
But does the weather actually have much of an impact? While evidence certainly suggests that it does in the US, a study published by Democratic Audit based on voter turnout in Sweden suggests that, as long as voting is convenient, weather does not affect turnout.
“Previous studies,” the authors write, “focused on the United States, a country where the costs associated with voting are high in a comparative perspective”.
In Sweden, where voting is easier – registration is not required and election day is on a weekend – the results are different. “Even when using datasets covering almost 150,000 persons and very detailed rain data, we do not find any meaningful effects of weather conditions” the report says.
Using the same metric of difficulty, voting is “easier” in the UK than America, but more difficult than Sweden. So what happens in the UK?
According to Stephen Fisher, a politics researcher at Oxford University, there is little correlation between good weather and voting patterns. He told the BBC in 2002 that data from the last 15 general elections showed no link.
Even if there was some relation between the weather and turnout, Fisher points out, a higher proportion of people voting by postal ballot would diminish the effect.
There is one caveat, however. “If you had a January snowstorm”, says John Curtice from the University of Strathclyde, “it would make a difference but for the most part the weather is mildly inclement at these times of year [when general elections are held]. So you might need to take a brolly with you or you might have sunshine but you won’t have a howling gale or snow or serious travel disruption.”
That last point is important. While rain might not put people off voting, difficulties getting to the polling station can. So a truly chaotic storm could affect a result.