No, the polls don't say Labour would lose 44 seats in 2020


First thing to note is the current point on the electoral cycle. Of course the leadership contest is likely to have an effect on the polls (although it would be terribly naive to think that will automatically be reversed should Corbyn win the contest), but the trend for opposition leaders has always been downward. After his first two months, Miliband’s Labour averaged 39% in the polls. Labour got 31% of the popular vote in 2015. Two months into Tony Blair’s leadership, Labour were on an astonishing 53% of the vote. But by Election Day, Labour could only manage a measly 43%. After the first two months of Corbyn’s reign, Labour averaged at 30%.

It could easily be that this poll, putting Labour at 27%, isn’t anywhere near as bad is it’s going to get. Looking at previous trends, it would be justifiable to expect Labour to finish the electoral cycle with somewhere between 20% and 25% of the vote. The Labour crisis hasn’t yet brought popular support below this level. By my calculation, that would make Labour’s losses closer to 70 than 40.

All of this is forgetting the giant elephant in the room. The boundary changes set to be in place by the 2020 election reduce the number of Westminster seats to 600 from 650, with an awful lot of those 50 Labour seats, while many safe seats across the board are being made a lot less safe. I suspect the Tories will have a lot less to worry about than Labour MPs like Lillian Greenwood. Her safe Labour seat in Nottingham will now take in enough countryside towns as to make it a theoretically Conservative-held marginal.

Tories caught up in these boundary changes aren’t in such a bad situation. Jane Ellison, the Tory MP for Battersea with a majority of 8,000, would find herself fighting a seat with a notional Labour majority of just under 3,000. If the national picture holds in London, 3,000 votes will be well within the swing to the Tories.

Taking this into account, Labour is on course for between 120 and 140 seats in a 600 seat parliament.

Of course, let’s not forget the Lib Dems. The ICM poll suggested they’d actually lose a seat. Under new boundaries they’d lose two or more. I suspect, from the huge regional differences in their vote share in council by-elections and in their new membership, theirs may be an even more complex picture. At least until we start getting constituency level polling in. Regardless, the Lib Dems seem to currently be stuck in the 5-15 seat window.

All of this can, and will, change the moment Labour splits. I don’t think there’s much doubt there’ll be some kind of split now, and all of these polls will then become irrelevant.