![](https://dl6pgk4f88hky.cloudfront.net/2021/06/gettyimages-1163571847-scaled.jpg)
There are two ways that Jo Swinson can hurt Boris Johnson in this election: doing very well at the election, and doing very badly. Do very well, and the Liberal Democrats gain far more seats at the Conservatives’ expense than they indirectly cost the Labour party in seats where the party has no hope of winning, making it impossible for him to go ahead with Brexit. Do very badly, and the Labour party recovers to its 2017 level, holds on to all or most of its seats in England and Wales, and the SNP retake enough seats from the Scottish Conservatives to shut Johnson out of office.
Thanks to our antiquated electoral system, it’s quite difficult to tell what “doing well” looks like until the evening of 12 December. One reason that the Liberal Democrat rank-and-file is excited about this election is they think that the party’s new pro-European voters are much more effectively located and concentrated than the voters it gained from the 1980s until it entered the coalition. Although the Liberal Democrats are still badly served by the United Kingdom’s electoral system, their 2017 electoral coalition was more efficient than their 2015 one, electing 12 MPs with fewer votes than they got in 2015, when they got eight. Many in that party think they can pull off similar gains thanks to a more concentrated Liberal Democrat core vote. But they could be wrong: the Liberal Democrat surge could be happening pretty evenly across the country in a way that has very little direct effect on the Liberal-Conservative battleground but a significant indirect effect on the Labour-Conservative battleground.
The polls are divided: essentially, of the companies doing regular polling, YouGov, Opinium, IpsosMori and Deltapoll show a Conservative lead that is probably big enough for gains at Labour’s expense to cancel out Liberal Democrat losses, while Survation and ComRes show a lead that probably isn’t. The pattern in Survation’s constituency polling shows remarkably well-distributed Liberal Democrat and Labour voting, with very few Liberal Democrat voters in the marginal Labour-held seat of Gedling, and very few Labour voters in the numerous Liberal Democrat marginals they have polled. Constituency polling is very difficult to do well and one reason to be dubious about it is the effectiveness of the Labour-Liberal Democrat vote in these polls suggests a very high-level of political engagement, and one reason why opinion polls have tended to get things wrong is that they have oversampled politically engaged people at the expense of the politically unengaged. (In 2015, politically engaged 2010 Liberal Democrat voters were more likely to have switched to Labour, while politically unengaged ones were more likely to have switched to the Conservatives.)
But we can say with reasonable certainty that at the point the Liberal Democrat vote is above 22 per cent, it starts to cause sufficiently large problems for the Conservatives that the indirect problems it causes for Labour are cancelled out, and that when it gets to below ten per cent, Labour start to do well enough that the Conservatives start to have real problems in the Conservative-Labour battleground. The good news for Boris Johnson is that the Liberal Democrats are not at 22 per cent in the polls and they are well above ten per cent.
The bad news is that we can’t say with any certainty what the net effect of the Liberal Democrats’ increased vote share is while they continue to poll at around 17 per cent. We know that there is an indirect benefit to the Conservatives and a direct cost to them. We just don’t know which of those is greater.
One of the ways that Jeremy Corbyn has a simpler job in this election is that his interests are pretty straightforward: the lower the Liberal Democrat voteshare, the better for him. He and Labour are doing a pretty good job of repeating and reinforcing the message that this election is a choice between him and Johnson.
The problem for Boris Johnson is that his best interests are not as clear: we can’t say exactly where the optimal Liberal Democrat performance is from a Conservative perspective but we can say that it is probably quite a small range.
But Johnson’s campaign is doing a ruthless and pretty efficient job of repeating the message that this campaign is a two-horse race between him and Jeremy Corbyn. He has even agreed to two one-on-one debates between himself and the Labour leader. It’s hard to see the benefit of these to Johnson: one-on-one debates are difficult for the incumbent, Corbyn is quite good at this format, and most importantly, Swinson’s absence from them underlines the message that Labour want to send, which is that this election is a choice between Labour and the Conservatives. Johnson may have cause to regret that.