Tories remain at rock-bottom
Mid Beds picture remains inconclusive
Huge swing projected in Tamworth
July saw three by-elections in one day. Whilst the results were good for Labour, confirming a disastrous position for the Conservatives, we didn't get the clean sweep to generate a decisive narrative. Three more are scheduled for October. Betfair odds favour Labour to win all three, but not by anything like the one-sided margin it did in the previous three.
Scottish race is a key General Election indicator
The first, on October 5th, will get the least attention from Westminster media but it could provide the most significant indicator. Rutherglen and Hamilton West is a straight fight between Labour and the SNP. Labour are rated 89% likely to win at odds of 1.121/8, but were even shorter earlier, at 1.041/25 (96%).
They really should be winning this comfortably, given wider polling trends. The SNP have fallen back considerably since Nicola Sturgeon's exit. A Labour Majority at the next general election - currently rated around 68% likely on our markets at 1.4740/85 - may very well hinge on them gaining at least a dozen seats in Scotland.
Rutherglen and Hamilton West is 54th on Labour's target list, and their fourth best chance of a pick-up in Scotland. They won it as recently as 2017, when trailing the SNP by 9.8% across Scotland as a whole. At the 2019 general election, that advantage increased sharply to 26.4% and SNP duly regained the seat.
Are SNP fighting back?
In the most recent poll, their national advantage was 11%. However in the previous eight, it had averaged a mere 2.5%.
The likeliest outcome is indeed a Labour gain, which would make very few waves in the news and be swiftly forgotten. However if that much better poll for the SNP is part of a wider trend, they could hold on and raise serious doubts about Labour winning an overall majority.
By-election betting projects Tories losing two safe seats
The two other by-elections are in Conservative-held English seats and will be held simultaneously on October 19th. Given an enormous advantage in 2019, losing either will be a hammer blow to Rishi Sunak. Both seats are way beyond the line Labour would need to win a majority.
Mid Bedfordshire - previewed recently here - has never elected anyone other than the Conservatives. Labour have never come close and it is anything but fertile territory. They require a 19.1% swing. Tamworth requires an even bigger swing at 21.3%, but this is a seat that Labour won as recently as 2010.
In both cases, the Tory chance is hampered by the reasons behind the by-elections. Nadine Dorries was the MP for Mid Beds, while Tamworth was represented by Chris Pincher - the former Deputy Chief Whip who resigned in the wake of a scandal in which he was accused of sexually assaulting two men.
Will Labour win Mid Beds? They're clear favs in latest odds
Since writing that Mid Beds preview, the odds have moved somewhat towards Labour. At odds of 2.447/5, They have now overtaken the Lib Dems at the head of the market, in the wake of this new poll.
It must be stressed that these numbers are highly inconclusive. Take it to the bank that 29% will not be the winning total. Also check out this thread regarding a focus group. The Mid Beds electorate are evidently largely undecided.
I'd suggest the likeliest next step in this race is a further shift from Lib Dems to Labour, as the narrative prevails that they are the choice of the 'tactical anti-Tory alliance'. We've seen that in by-election after by-election but be wary. Whereas there is a long correlation between Labour and Lib Dem shares in most constituencies, they have consistently remained separate tribes in this one.
Don't rule out the Tories either. While 29% would be a truly awful result for them in such a core constituency, 35% isn't a huge ask and that could be the winning total. If pushed at this stage to predict the result right now, I'd go Labour 36, Conservative 33, Lib Dems 14, Gareth Mackey 10, Reform 5.
Enormous swing is realistic in Tamworth by-election
Tamworth is a much clearer, two-horse race, with Labour already backed heavily down to 1.331/3. The required swing of 21.3% is indeed huge, but Labour managed 23.6% in Selby and Ainsty.
When weighing these required swings, I think it is important to consider swings over the previous decade. Certain type of seats - older, whiter, fewer graduates and public sector workers - swung much further to the Tories than the median, as the electorate re-aligned along those cultural and demographic lines.
Thus the Labour vote has already been squeezed to the max, leaving much greater potential to grow now the national picture has transformed in their favour. Whereas in a seat like Uxbridge, where they managed a mere 6.7% swing on the same day as Selby and Ainsty, they had far less room to grow because the demographic re-alignment had previously worked in their favour.
Tamworth has proved a General Election bellwether
Tamworth is precisely the sort of seat where Labour have plunged to all-time lows during that era. It used to be considered a bellwether. Indeed if also including its predecessor, South East Staffordshire, it voted for the winning party in every general election after creation in 1983 and swung heavily towards Labour in a 1996 by-election, just before Tony Blair came to power.
Plus here, even more than the Dorries example, the former MP's disgrace is likely to be a factor. The Pincher scandal was big news. If the Tories are ripe for a mid-term protest beating anywhere, Tamworth is a very likely candidate.
The national picture remains as grim, and apparently set, for the government as it did in July. There is no evidence as yet of a comeback and multiple polls have them around 25%. With widespread tactical voting expected, this level of support represents an existential threat.
Party conference season will likely add to a constantly awful news cycle. For my money, Tamworth will add to it. Rishi Sunak's best hope lies in Mid Beds. Retaining it would not indicate any sort of comeback, but it might at least neuter the worst headlines.