Is there any way there might be an early Westminster general election before 2028?

It's fair to say that an early general election would probably be in the overwhelming best interests of the SNP and the independence movement.  The polls do not show the SNP on a stellar vote share, but that doesn't matter under first-past-the-post as long as you're in first place and each other party is far enough behind you, and that is exactly the scenario that has arisen.  An early election would very likely see the SNP regain a majority, perhaps even a landslide majority, of Scottish seats at Westminster, and the 2024 election would be left looking like an unimportant aberration.

The snag is, of course, that an early election is extraordinarily unlikely.  There's no incentive for Keir Starmer to call one, in fact there's every conceivable disincentive.  And his hand can't be forced with a vote of no confidence, because he has an overwhelming majority of Commons seats (albeit won on only 34% of the popular vote).  We quite rightly scoffed at the Trump allies in the US who apparently spent January and February fantasising about how they were going to contrive a way of bringing the Labour government down early.

Nevertheless, very strange things do happen sometimes - nobody imagined in the spring of 2017 that Theresa May was only a few weeks away from losing her parliamentary majority, because she didn't have to call an election for another three years.  So just from a purely speculative point of view, I've been pondering whether there might be some sort of way an election could happen before, say, 2028 (and some would argue that even 2028 would count as "early").

I think almost certainly the key to it would be Keir Starmer's early departure as Labour leader.  That would have two effects - first of all, it would give the green light to the London media to start chipping away at the new Prime Minister's legitimacy, ie. they'd say that he or she lacks any 'personal mandate'.  Secondly, and far more importantly, it would open up the possibility of the Labour membership selecting a leader who a significant percentage of Labour MPs cannot live with.

Imagine, for example, that Jeremy Corbyn was currently Labour Prime Minister with 400 MPs.  That would plainly not be a stable situation in which we could be sure that the parliament would survive its full term, because many, and perhaps most, of those 400 MPs would regard themselves as informally part of the opposition to the Corbyn government.  Now, of course Starmer's replacement isn't going to be a Corbynite, the rules have been stitched up to ensure that can't happen.  

But what if it was someone from the soft left?  What if it was Angela Rayner, for example?  Is it just conceivable that the Labour parliamentary party has moved so far to the right that there are several dozen MPs who wouldn't even be able to stomach someone like Rayner?  There might also be a class element to it - there are a fair number of posh boy (and posh girl) Labour MPs who would cringe every time they hear their own Prime Minister speaking with that accent.  Suppose the right-wingers organised and went all-out to bring down Prime Minister Rayner, and when they failed went for the Plan B of setting up a Change UK style organisation, but on a much bigger scale?  That might just possibly be enough to bring down the government and trigger an early election.  It's an extreme long shot, but not totally impossible.

As you might have seen, there's been a lot of chatter in the media in recent days about Rayner supposedly already being on manoeuvres, and one Labour source was quoted as saying the leadership battle would undoubtedly be a straight fight between Rayner and Wes Streeting.  If that's true, punters at the betting exchanges don't seem to have twigged yet.  Streeting is the weak favourite, but is only priced at 7.8, while Rayner is only priced at 8.8, implying a strong probability that the next leader will be someone else entirely.  Rayner's odds have shortened since I last checked, but I still think she looks like value.