Who’ll win the Brecon and Radnorshire by-election? We’ll know in a few hours but it’s worth remembering the following: if I told you that, nine years into the life of a government, with the economy looking a little peaky, a by-election was being held in a constituency which an opposition party had a) held and b) still held on the same boundaries at a devolved level, you wouldn’t even think about it.
Then if you added that, in addition, the by-election had been triggered because the sitting MP had been convicted of expenses fraud – albeit in a case in which the court conceded that the fraud had been unwitting – and that 10,000 constituents had signed a petition calling on him to be recalled, and that the governing party had nominated that same MP as their candidate, it feels unlikely that you would bother staying up for the result.
The only complicating factor is that the by-election coincides with the governing party getting a bounce from having a new leader in place, but even then, you’d assume that all things being equal, the challenging party is likely well organised enough to have build up a big lead thanks to postal votes to be well-insulated against a late swing. We should also note that, if the polls are right, we’d expect the Liberal Democrats to win this seat even after Boris Johnson’s bounce in the polls at a general election, and opposition parties tend to overperform in by-elections compared to at general elections.
All of which is to say that if the Conservatives have held the seat it would be a coup for the party in general and Boris Johnson in particular. It would suggest that the polls are actually underestimating the scale of his support, or the general strength of the Brexit party in general. But, of course, it could equally just be that something has gone badly wrong with the Liberal Democrat campaign on the ground or that it’s all noise.
Whatever the result, we should expect the Labour vote to be way down – even in years of overall Labour strength, like 1993, when they gained 111 councillors in the local elections and were polling an average of 42 per cent, their vote was badly squeezed in Liberal-Conservative battleground seats, as we’d sort of expect. In 1997, in Winchester, their vote collapsed in a Liberal Democrat-Conservative by-election (the Liberal Democrats won). The only exceptions are the very peak of Tony Blair’s popularity in the run-up to the 1997 election, when the Liberal Democrats made gains from the Conservative party and the Labour vote also went up.
So all we might learn tonight about Labour is that the Jeremy Corbyn of 2019 is not as popular as the Tony Blair of 1995 – which frankly is not suprising, or something we can describe as a lesson in a meaningful sense. It would be truly remarkable if the Labour vote held up and would suggest that something was either going very right for Labour or very wrong for the Liberal Democrats, whether nationally or simply in Brecon and Radnorshire.
What if the expected happens? Well, it would further suggest that the story in the polls since Johnson became Conservative leader, that Johnson has decreased the flow of 2017 Conservative voters to the Brexit party but increased the flow to the Liberal Democrats.
But we still won’t know very much – if the result looks anything like what we expect, well, we already have a pretty good idea that the price of a Johnson leadership is that the Tories are going to have a harder time in territory where the Liberal Democrats are strong, whether historically or more importantly in the local and European elections, and anywhere where Remainers cluster in greater numbers than the national average.
What Conservative MPs hope is that the product – doing better in areas where there are more Remainers than the national average and in the Conservative-Labour battlegrounds in the Midlands, parts of Wales and the North – is worth that price, that what they lose to the Liberal Democrats and the SNP is cancelled or even better exceeded by what they gain from Labour.
And we will be no closer to knowing if that calculation is wrong tonight, whatever the result – though we might get a clue that it is right.