New Times,
New Thinking.

  1. Politics
  2. The Staggers
30 April 2021

What would be a good result for Labour in the Hartlepool by-election?

A defeat in the by-election or a narrow victory would both be worrying for Keir Starmer’s party.

By Stephen Bush

Predicting the Hartlepool by-election is ultimately about deciding what type of constituency you think the seat is. If you think Hartlepool should be considered a “normal” Labour seat, then the answer is pretty easy: not even YouGov, who have consistently shown the largest Conservative poll leads, suggests that the Tories are in a position to gain a parliamentary seat off the opposition. (The only two times that has happened in the past half-century, the government was consistently polling more than 15 points ahead of the opposition.) 

But Hartlepool is an unusual constituency: one in which the Brexit Party did much better than in almost any other Labour-held seat (the Brexit Party did not stand in constituencies the Conservatives already held in 2019). So the other analysis of the constituency holds that to understand the balance of forces in Hartlepool you need to start by reallocating the Brexit Party vote between the two parties – say at a two-to-one margin to the benefit of the Conservatives. Hartlepool shouldn’t be understood as a “normal” by-election in opposition territory, but one held by the government: a reasonable adjustment would say Hartlepool “starts” with a Tory majority of, say, 5 per cent. And just as we can say very confidently that the Conservatives are not polling well enough to win a by-election in opposition territory, we can say pretty confidently that Labour isn’t polling well enough to win Conservative-held marginals of any description.

Either analysis of the constituency is, I think, pretty reasonable. (As is the “third way” that Hartlepool is a strange constituency where independents have long been unusually successful in council elections, where Peter Mandelson ran ahead of the Labour Party nationally on each occasion, and where the Conservatives have done poorly compared to both the national and regional picture.) My starting point is that the quarter of Hartlepool voters who voted for the Brexit Party in December 2019 knew what they were voting for, and that their conscious choice to vote for neither the Conservatives nor Labour means that we should treat this as a normal by-election in an opposition constituency until convincing proof emerges otherwise. But I could be wrong.

If I am wrong, I suspect that the line to take among Labour talking heads and supporters of the party’s leadership will be that this by-election, as well as coinciding with the Tories’ vaccine bounce, is an odd seat and should be treated and analysed as a Conservative hold rather than a Conservative gain.

I think this will be a good example of something being “incidentally correct”: in that while this will be true, I don’t think it matters at all. While those two competing analyses of Hartlepool are useful if you’re trying to make money betting on the Hartlepool by-election, they aren’t really meaningfully distinct categories if you are analysing what the by-election means about the current political position of the two parties.

If Hartlepool should be seen as a Labour-held constituency with a reduced but still healthy 8-point majority, then Labour ought to win it comfortably if they are currently in a position to win a parliamentary majority. But if Hartlepool is “really” a lot more like the Conservative seats with a small majority of say, 5 points, then Labour ought to win it comfortably if they are currently in a position to win a parliamentary majority. In fact, even if you go for the most plausible pro-Conservative reassignment of the Brexit Party vote, the seat is still a constituency Labour ought to be winning if they are on course to increase the number of seats they hold, let alone to win an election or form a government.

For the Hartlepool result to be “good” for Labour, the result should not be close: it should not be close if we count it as “a seat held by the opposition” or if we count it as “a seat held by the governing party with a majority of less than 5 points”. It’s not winning or losing Hartlepool that tells us much about Labour’s current electoral prospects: but the margin, which should, if all is rosy in the Labour garden, be large, however you analyse the seat.

Give a gift subscription to the New Statesman this Christmas from just £49

Frequently asked questions

You can’t possibly think Labour will comfortably win this seat? 

These articles are not predictions. They are an attempt to set out reasonable yardsticks.

There are three other elections taking place in Hartlepool on the same day. Does this change your thinking at all? 

Yes and no. If it turns out that Labour have won Hartlepool in the council election, won the contiguous wards in the police and crime commissioner and the metro-mayoral election, but lost the constituency, then I would conclude that the party was in an OK place electorally but had suffered through having a poor candidate – which would, of course, reflect poorly on the Labour leadership in other ways, but might not tell us a great deal about the party’s current electoral position.

If Labour does poorly because Ben Houchen, the Conservative metro-mayoral candidate, carries all before him and helps elect Jill Mortimer as the town’s MP, then it might, again, suggest that we can’t learn all that much from this constituency alone. 

But of course, neither of those results would be positive for Labour: successful parties ought to select good candidates and win metro-mayoralties, so I don’t think we need to be all that worried either way.

Labour had a better record in by-elections in the 2017-19 parliament, which ended in catastrophic defeat, than the 2015-17 one, which ended in a hung parliament. Harold Wilson, who won four general elections, had a significantly worse by-election record than Neil Kinnock, who lost two. Seriously, isn’t this is a pointless exercise?

Yeah, a bit. By-elections are “statistically noisy”: there’s a lot going on and a lot can change. I think they are significantly less revealing than local elections – which tell us all sorts about regional patterns of support – not least because there are fewer of them. But while I think they are less useful yardsticks than local elections, they are still useful.

Content from our partners
Putting citizen experience at the heart of AI-driven public services
Skills policy and industrial strategies must be joined up
How the UK can lead the transition to net zero